Category Archives: Uncategorized

Off-Grid Power Architectures For Remote And Edge Data Centers In Energy-Constrained Environments: A Technical, Economic, And Resilience-Centered Research Review

Authors: Samuel N Nimaful, Augustine Hanyabui, Joel Holison

Abstract: Remote and edge data centers are increasingly deployed in locations where grid power is unavailable, unreliable, capacity-constrained, or prohibitively expensive. In these contexts, “off-grid” practicalities are less about complete electrical isolation than about assured energy autonomy: the ability to maintain service-level objectives (SLOs) and critical uptime during prolonged power interruptions, fuel supply disruptions, and extreme environmental conditions. Achieving this autonomy requires power architectures that integrate dispatchable generation (diesel or gas gensets and/or fuel cells), variable renewable energy (VRE) resources (solar PV, wind, and in some locations hydro), energy storage (UPS and BESS), robust power electronics (including grid-forming inverter-based resources), and supervisory energy management systems (EMS) that co-optimize reliability, cost, and emissions. This paper addresses the research problem: How can off-grid power systems for remote and edge data centers be architected and operated to meet high availability targets under energy constraints while minimizing lifecycle cost and carbon emissions? It synthesizes standards-body guidance, government laboratory research, recent peer-reviewed literature (2016–2026), and vendor technical documents into design patterns, a quantitative comparative model, and actionable deployment guidance. Key findings are as follows. First, microgrids structured around a formal controller specification (e.g., microgrid controller functional requirements in IEEE microgrid-controller standards) provide an engineering basis for predictable islanded operation, black start, and coordinated dispatch across distributed energy resources (DER). [1] Second, hybridization is the dominant pathway for energy-constrained environments: diesel-only designs are simple but are exposed to fuel logistics, price volatility, and emissions; adding renewables and storage materially reduces fuel burn and can improve resilience by reducing the frequency and severity of fuel-delivery dependency—an especially salient risk in remote microgrids where delivered diesel electricity can be extremely costly. [2] Third, for off-grid stability and fast contingency response, inverter-based resources and their protection/control behaviors (grid-forming operation, current limiting, and black-start behavior) are increasingly central, especially as renewable penetration rises. [3] Fourth, safety and compliance for stationary storage (e.g., fire and thermal-runaway propagation testing and installation codes) are not peripheral—they shape siting, enclosure design, and permitting timelines and thus can dominate schedule risk. [4] Quantitatively, a parametric cost-and-carbon model demonstrates that (i) LCOE and emissions are strongly driven by delivered fuel price and renewable fraction, and (ii) heavier “soft costs” and integration overhead penalize very small deployments unless modularized and standardized. Using published CAPEX/O&M baselines for PV, wind, BESS, and gensets, and modeling three load scenarios (low/medium/high) with sensitivity to delivered diesel price, the modeled LCOE ranges from roughly

.20–

.20–

.70/kWh depending on architecture and fuel price, while carbon intensity ranges from ~0.26–0.74 kg CO₂/kWh as renewable delivered share rises from ~0% to ~65%. [5] Finally, three geographically diverse real-world examples illustrate the range of viable approaches: a gas-generator solution for a large Lagos data center where grid reliability was insufficient; a fuel-cell-powered containerized edge data center integrated with district heating in northern Sweden; and an Alaska edge deployment co-located with hydropower and backed by advanced microgrid modernization efforts—each reflecting different constraints and resource endowments. [6]

DOI: https://doi.org/10.5281/zenodo.19414625

AI-Based Online Proctoring System For Secure And Scalable Remote Examinations

Authors: Mayur Patil, Kunal Viroje, Harsh Waingankar, Dr. Vivek Khalane, Dr. Vaibhav Narawade

Abstract: Online examinations have become a common part of modern education, especially with the growth of remote learning platforms. However, maintaining fairness and preventing malpractice in such environments remains a major challenge. In this work, we present an AI-based online proctoring system designed to monitor candidates during examinations using real-time video and audio analysis. The system combines face recognition, gaze tracking, head pose estimation, and audio monitoring to detect suspicious activities such as impersonation, presence of multiple individuals, and abnormal behavior. During our testing across multiple sessions and varying environmental conditions, we observed that the system achieved an overall detection accuracy of approximately 92.6% while maintaining real-time performance of 20–30 frames per second. The proposed system reduces dependency on human invigilators and provides a scalable solution for large-scale online examinations.

Synthesis, Spectroscopic, And Biological Studies Of Complexes Of Unsymmetrical Thiosemicarbohydrazide Ligand

Authors: Tanhaji Walunj, Madhukar Badgujar

Abstract: A new unsymmetrical p-fluorobenzaldehyde derivative of α-benzilmonoximethiosemicarbohydrazide (HBMTSpFB) ligand is prepared via condensation of α-benzilmonoximethiosemicarbohydrazide and p-fluorobenzaldehyde in the 1:1 ratio. Metal complexes of Fe(II), Co(II), Cu(II), Zn(II), Hg(II) and Ni(II) have been prepared. These prepared compounds were characterized by physicochemical study, PMR, FT(IR), electronic absorption, and magnetic moment, and the purity of the HBMTSpFB ligand was analyzed by thin layer chromatography study. All prepared compounds are color-solid, air-stable, and soluble in common organic solvents. On the basis of elemental analysis metal to ligand and stoichiometry is 1:2 ratio for all complexes. Comparison of the FT(IR) spectra of the HBMTSpFB ligand and its trivalent metal complexes confirm that the HBMTSpFB ligand is a monobasic, tridentate ligand towards the central trivalent metal ion with an ONS and sequence.

DOI: https://doi.org/10.5281/zenodo.19415327

Self-Generating Hybrid Aluminum-Assisted Green Hydrogen System

Authors: Suren kumar Selvamani

Abstract: This work presents a hybrid aluminium-assisted hydrogen generation system utilizing waste aluminium feedstock for continuous hydrogen production through a combination of chemical reaction and electrolysis. Aluminium scrap is processed into fine particles and reacted with water in the presence of a catalyst to generate hydrogen. The system integrates a secondary electrolysis unit to extract additional hydrogen from residual water, thereby improving overall efficiency. A catalyst regeneration loop is incorporated to enable repeated use of catalytic material, while aluminium is consumed as an energy carrier and converted into aluminium oxide. The system is designed for decentralized, on-demand hydrogen generation, particularly suited for remote, off-grid, and waste-to-energy applications.

Intelligent Phishing Website Detection Using Machine Learning For Secure Online Systems

Authors: Sagar Kumar, Harish Dutt Sharma, Ram Bhawan Singh

Abstract: Phishing attacks have emerged as one of the most significant cybersecurity threats, targeting users by creating fraudulent websites that mimic legitimate platforms to steal sensitive information. Traditional rule-based and blacklist-based detection techniques are often ineffective against newly generated phishing websites. This paper proposes a machine learning-based phishing website detection system that utilizes multiple classification algorithms to identify malicious URLs. The system extracts various URL-based and domain-based features such as URL length, presence of special characters, domain age, and HTTPS usage. Machine learning models including Support Vector Machine (SVM), Random Forest (RF), and Logistic Regression (LR) are evaluated. Experimental results demonstrate that the proposed approach achieves high accuracy and outperforms traditional detection methods.

DOI: https://doi.org/10.5281/zenodo.19415751

Synthesis And Characterization Of Several Transition Metal Complexes Derived From α-benzilmonoximethiosemicarbohydrazide And M-chlorobenzaldehyde.

Authors: Sandip Thube, Dr. M. A. Badgujar

Abstract: Several complexes derived from thiosemicarbohydrazide, specifically α-benzilmonoximethiosemicarbohydrazide-m-chlorobenzaldehyde (HBMTSmCB) and its complexes with Fe(II), Ni(II), Cu(II), and Co(II), have been synthesized and meticulously characterized. The characterization employed a range of analytical techniques, including elemental analysis, conductivity measurements, and magnetic susceptibility assessments. Spectroscopic methods such as Proton Magnetic Resonance (PMR), Fourier Transform Infrared (FTIR) spectroscopy, and electronic absorption spectra were also utilized to elucidate the structural and bonding characteristics of these complexes. It was determined that all trivalent metal complexes synthesized exhibit octahedral geometries.

DOI: https://doi.org/10.5281/zenodo.19415792

Explicit Dynamic Frontal Crash Test Analysis Of FSAE Roll Cage Using AISI 4130 And Docol R8 Steel

Authors: Sagar Nadavati, Jeffrey Joe, Janakiraman

Abstract: This study investigates the crashworthiness performance of a FSAE roll cage subjected to frontal impact using explicit dynamic simulation. Two high-strength materials, AISI 4130 chromoly steel and Docol R8 advanced high- strength steel, were evaluated. The roll cage geometry was modelled using SolidWorks and imported into ANSYS Explicit Dynamics for frontal crash simulation at an impact velocity of 8 m/s against a rigid wall boundary condition. Key performance indicators such as total deformation, equivalent von-Mises stress distribution, plastic strain, and energy absorption characteristics were analysed. A comparative study between both materials was conducted to determine structural safety performance and weight optimization potential. Results indicate that Docol R8 provides improved strength-to-weight performance compared to AISI 4130, demonstrating its suitability as an alternative roll cage material for Formula Student vehicles.

DOI: https://doi.org/10.5281/zenodo.19415920

Sketch Rush: A Real-Time Digital Pictionary Experience

Authors: P.Vijay, S.Manikanth, S.Chaitanya, V.Ramya

Abstract: In an era dominated by digital communication, traditional social games that rely on physical presence and non-verbal interaction face the risk of obsolescence. Games like Pictionary, which thrive on creativity, quick thinking, and shared laughter, are often difficult to replicate in a virtual environment without losing their core essence. To address this, we present Sketch Rush: A Real- Time Digital Pictionary Experience, a web-based multiplayer game that faithfully recreates the excitement and social dynamics of the classic drawing and guessing game. Sketch Rush leverages modern web technologies to provide a seamless, interactive platform where players can connect, create, and compete in real-time. Sketch Rush is designed not merely as a digital adaptation but as an enhanced, accessible version of the original game. It addresses the limitations of physical Pictionary—such as the need for physical drawing tools, proximity of players, and manual scorekeeping—by automating these processes within an intuitive digital interface. The system comprises two primary modules: a real-time drawing canvas with a rich set of tools for the “Artist,” and a dynamic chat interface for the “Guessers.” The core game logic, powered by a Node.js backend and WebSocket communication, ensures low-latency synchronization of drawings, guesses, and game states across all connected clients. Preliminary user testing with a cohort of 40 participants has shown that Sketch Rush successfully captures the engaging and collaborative spirit of the original game. Feedback highlighted the platform’s intuitive interface, the responsiveness of the real-time features, and its effectiveness in fostering social connection, even among geographically dispersed players. Users reported a high degree of satisfaction, with average System Usability Scale (SUS) scores of 85.6, indicating excellent usability. In essence, Sketch Rush reimagines a beloved social game for the digital age. It transcends the limitations of physical location, offering a platform that is not only functional but also fun, engaging, and socially enriching. By combining intuitive design with robust real-time technology, Sketch Rush provides a compelling case for the successful digital transformation of traditional social experiences.

DOI: https://doi.org/10.5281/zenodo.19416157

A Study On Strategies Plan For Inclusive Digital Rural Development

Authors: Harish M, Dr.M.D.Chinnu

Abstract: This study focuses on strategies for inclusive digital rural development, aiming to bridge the digital gap between rural and urban areas. It examines the availability of digital infrastructure, access to technology, and the level of digital literacy in rural communities. The study also explores the challenges faced by rural people in using digital services, especially women, elderly, and marginalized groups. It highlights the role of digital tools in improving education, healthcare, financial services, and rural livelihoods. Both primary and secondary data are used to understand the current situation and identify gaps. The research emphasizes the importance of government support, policy planning, and skill development programs. Based on the findings, effective strategies are suggested to promote inclusive and sustainable digital growth. Overall, the study aims to support balanced development and empower rural communities through digital inclusion.

Food Safety, Animal Health, And Environmental Sustainability: A Policy Integration Model

Authors: Dr. Geetika

Abstract: The inter-linkages between environmental contamination, animal health, and food safety have emerged as critical concerns in the context of rapid industrialization and agricultural intensification. This study develops a policy integration model grounded in the One Health framework, using empirical evidence from Haryana, India. Heavy metals and pesticide residues originating from industrial and agricultural activities were traced across soil, water, livestock feed, and milk, demonstrating systemic transfer through the food chain. Health risk assessment indices, including Estimated Daily Intake (EDI), Hazard Quotient (HQ), and Cancer Risk (CR), indicate potential human health implications. The findings highlight the inadequacy of fragmented governance systems and propose an integrated, multi-sectoral policy model aligned with global sustainability goals. This research contributes to bridging the gap between environmental science and policy design, offering actionable insights for developing economies.

Smart Crypt-Based Secure Storage And Fine-Grained Sharing Of Time-Series Data Streams In Industrial Internet Of Things

Authors: Mrs.K.Sham Sri, Repuri P S S Chaitanya, Karibandi Manasa, Indraganti Sai Teja, Dulam Shiva, Kona Venkata Satya Sai Kumar

Abstract: The rapid growth of the Industrial Internet of Things (IIoT) has led to the continuous generation of large volumes of time-series data from sensors and industrial devices. These data streams are commonly stored and processed in cloud platforms to enable scalability, remote monitoring, and advanced analytics. However, storing sensitive industrial data in cloud environments introduces significant privacy and security risks, including unauthorized access and data breaches. To address these challenges, a secure data storage and sharing framework for time-series data streams in IIoT environments is proposed. The system employs a symmetric homomorphic encryption technique that enables analytics to be performed directly on encrypted data without revealing the original information. Additionally, the framework introduces fine-grained access control mechanisms that allow data owners to selectively share encrypted data streams with authorized third-party services. A verification mechanism based on message authentication ensures data integrity and authenticity during data processing and sharing. The proposed SmartCrypt-based approach enhances data confidentiality while maintaining efficient query processing and analytics capabilities. Experimental analysis demonstrates that the system improves query performance and throughput compared to existing encrypted data stream processing solutions, making it suitable for secure and scalable IIoT data management.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.162

Intelligent Traffic Signal Optimization Using Image Processing And Canny Edge Detection For Density-Based Traffic Management

Authors: Mr.KVV. SubbaRao, Neyigapula Jayakrishna, Meesala Venkata Sai Gnana Prakash, Pinninti Lakshmi Prasanna, Kallepalli Ramesh

Abstract: Traffic congestion has become a major challenge in urban transportation systems due to the increasing number of vehicles on roads. Conventional traffic signal systems generally operate on fixed timers, which often results in inefficient traffic management and unnecessary waiting time at intersections. To address this issue, an intelligent traffic control system based on image processing techniques is proposed. The system captures real-time traffic images using surveillance cameras and processes them to estimate vehicle density. The captured images undergo preprocessing operations such as grayscale conversion and noise reduction before applying the Canny edge detection algorithm to identify vehicle edges. The density of vehicles is determined by calculating the number of edge pixels in the processed image and comparing them with a reference image. In addition, the You Only Look Once (YOLO) object detection algorithm is used to identify emergency vehicles such as ambulances and provide them with priority signal allocation. Based on the estimated traffic density, the system dynamically adjusts traffic signal duration for each lane. The proposed approach improves traffic flow efficiency, reduces waiting time, and enhances emergency vehicle movement at intersections. This intelligent system can serve as a practical solution for modern smart city traffic management.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.163

Quantum Machine Learning Framework For Image Classification Using ResNet-Based Feature Extraction And QSVM

Authors: Ms.A.Harini, Battina Sai Mounika, Kondeti Sushan Niharika, Mummidi Rajesh, Sodasani Hari Veera Narasimha Manikanta, Kalla Vinod

Abstract: Image classification has become a fundamental task in computer vision with applications in areas such as medical imaging, agriculture, environmental monitoring, and automated surveillance. Traditional machine learning techniques have achieved reasonable performance in classification tasks; however, they often struggle when dealing with high-dimensional and complex image datasets. Deep learning models, particularly Convolutional Neural Networks (CNNs), have significantly improved image classification performance by automatically learning hierarchical feature representations. Despite these advancements, classical deep learning models may still face challenges related to computational complexity and large-scale data processing.In recent years, quantum machine learning has emerged as a promising paradigm that combines principles of quantum computing with classical machine learning techniques to enhance computational efficiency and model performance. This study proposes a hybrid quantum–classical framework for image classification that integrates a deep residual network (ResNet-50) with a Quantum Support Vector Machine (QSVM). The ResNet-50 model is employed as a feature extraction mechanism to capture high-level visual representations from image data. The extracted features are then reduced in dimensionality using Principal Component Analysis (PCA) to simplify the feature space and improve computational efficiency.The reduced feature vectors are subsequently classified using a QSVM model that utilizes quantum feature maps to encode classical data into quantum states. Various quantum feature maps are explored to evaluate their impact on classification performance. Experimental results demonstrate that the hybrid quantum–classical approach achieves higher classification accuracy compared to conventional machine learning models such as Support Vector Machines and Random Forest classifiers. The proposed framework highlights the potential of combining classical deep learning architectures with quantum machine learning algorithms to address complex image classification challenges. This hybrid approach provides an efficient and scalable solution for advanced image analysis tasks and demonstrates the growing potential of quantum computing in artificial intelligence applications.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.164

AI-Based Computer Vision System For Intelligent Rice Quality Classification Using Deep Learning And XAI

Authors: Mrs.P.Lakshmi Satya, Dadala Aksha, Pandrangi Sri Venkata Arya, Akula Raja, Pithani Hemalatha, Thota Venkata Subha Santosh

Abstract: Rice quality assessment plays a crucial role in the food industry as it directly affects consumer satisfaction, market value, and food safety. Traditional rice inspection methods rely mainly on manual observation and mechanical tools, which are time-consuming, labour-intensive, and prone to human error. To address these limitations, this study proposes an intelligent computer vision framework for automated rice quality assessment using deep learning and explainable artificial intelligence techniques. The system captures high-resolution images of rice grains and applies image preprocessing techniques such as grayscale conversion, edge detection, and segmentation to extract important visual features. Deep learning models, including VGG16 and ResNet50, are used to learn complex feature representations and classify rice grains based on their physical attributes such as size, shape, texture, and colour. To improve transparency and interpretability of the model predictions, Explainable AI (XAI) techniques such as Local Interpretable Model-Agnostic Explanations (LIME) and Gradient-weighted Class Activation Mapping (Grad-CAM) are integrated into the framework. Experimental results demonstrate that the proposed approach significantly improves classification accuracy and reliability compared to traditional inspection methods. The developed system provides an efficient, scalable, and automated solution for rice quality evaluation in agricultural and food processing industries.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.165

An Intelligent Wastewater Pollution Detection Framework Using Deep Learning And Sensor-Based Environmental Monitoring

Authors: Mr.G.Vijay Kumar, Pathi Krishna Kanth, Srikakolapu Chandi Mohana Manjusha, Palacharla Vidhatri, Makineedi Hari Gangadhar Satya Sairam, Bathula James

Abstract: Water pollution has become a major environmental concern due to the increasing discharge of industrial and domestic contaminants into wastewater systems. Continuous monitoring of wastewater quality is essential to detect harmful pollutants and prevent environmental damage. This study proposes an intelligent wastewater pollution detection system that integrates low-cost multisensor technology with deep learning techniques. The system collects environmental data using multiple sensors capable of measuring chemical characteristics present in wastewater. The acquired sensor data is pre-processed and transformed into structured textual representations, enabling advanced machine learning models to analyse patterns associated with different pollutants. A deep learning model based on transformer architecture is then employed to classify and identify contaminants present in the wastewater. The proposed approach improves detection accuracy while maintaining computational efficiency. Experimental evaluation demonstrates that the system achieves higher classification performance compared to conventional machine learning methods. The developed framework provides a cost-effective and scalable solution for real-time wastewater monitoring and environmental protection. Future improvements may include integration with IoT-based monitoring platforms and deployment in large-scale environmental monitoring systems.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.166

Machine Learning–Based Heart Disease Prediction System For Early Clinical Diagnosis

Authors: Dr.K.ChandraSekhar, Sathi Sudharshan Reddy, Anakapalli Bhargavi, Ulli Sri Satyasai Ramcharan Teja, Gubbala Y V Ganesh Kumar, Kakara Vivek

Abstract: Heart disease remains one of the leading causes of death worldwide, making early detection and accurate diagnosis essential for improving patient outcomes. Traditional diagnostic approaches often rely on clinical examinations and expensive medical tests, which may not always be accessible in all healthcare environments. In this research, we explore the use of machine learning techniques to develop an intelligent system for predicting the presence of heart disease using clinical parameters such as age, gender, blood pressure, cholesterol level, and heart rate. The dataset used in this study contains labelled medical records that are pre-processed, balanced, and divided into training and testing sets to ensure reliable model evaluation. Several supervised machine learning algorithms, including Logistic Regression, Support Vector Machines, Naïve Bayes, Decision Trees, K-Nearest Neighbors, and Linear Discriminant Analysis, are implemented and compared to identify the most effective model for heart disease diagnosis. Feature selection techniques are applied to determine the most influential clinical attributes contributing to disease prediction. To evaluate model performance, we employ a 5-fold cross-validation approach along with evaluation metrics such as accuracy, precision, recall, F1-score, and area under the ROC curve (AUC).Experimental results demonstrate that the Logistic Regression and Linear Discriminant models achieve the highest prediction accuracy, showing strong capability in identifying heart disease risk from clinical data. In addition, the integration of optimized feature selection methods improves the overall diagnostic performance while reducing computational complexity. The proposed machine learning framework provides an effective and scalable approach for supporting early heart disease detection and assisting healthcare professionals in clinical decision-making.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.167

Intelligent Toxic Comment Detection Using Machine Learning And Natural Language Processing Techniques

Authors: Dr.S.Suresh, Namala Sireesha, Shaik Davud, Tirumani Bhanu Shankar Satyanarayana, Kada Rama Satya Pavan, Kala Tirumala Venkata Sai Teja

Abstract: The rapid expansion of social media platforms and online communication systems has significantly increased the amount of user-generated content on the internet. While these platforms enable people to share ideas and communicate freely, they also expose users to harmful content such as hate speech, offensive language, cyberbullying, and abusive comments. Toxic comments not only affect healthy online discussions but also create negative psychological and social impacts on individuals. Therefore, developing automated systems capable of detecting and filtering toxic comments has become an important research problem in natural language processing and online content moderation. This study presents an intelligent framework for detecting toxic comments using machine learning and natural language processing techniques. The proposed system analyses textual data collected from online platforms and classifies comments into toxic and non-toxic categories. Various preprocessing techniques such as tokenization, stop-word removal, text normalization, and lemmatization are applied to clean and prepare the dataset for model training. Feature extraction methods including Term Frequency–Inverse Document Frequency (TF-IDF) and word embedding techniques are used to transform textual data into numerical representations suitable for machine learning models. Several machine learning and deep learning algorithms, including Naive Bayes, Support Vector Machines (SVM), Logistic Regression, and Convolutional Neural Networks (CNN), are implemented and compared to determine the most effective model for toxic comment classification. The models are evaluated using standard performance metrics such as accuracy, precision, recall, and F1-score. Experimental results indicate that deep learning models, particularly CNN-based architectures, achieve higher classification accuracy and better performance in detecting complex toxic language patterns. The proposed system can assist online platforms in automatically identifying harmful content and maintaining safer digital communication environments. By integrating machine learning techniques with advanced natural language processing methods, the framework contributes to improving online content moderation and promoting respectful interactions in digital communities.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.168

Quantum Computing–Driven Framework For Cryptocurrency Market Analysis And Price Forecasting

Authors: Dr. Manjula Devarakonda Venkata, Jagilinki Hemanjali, Datla Siva Rama Raju, Karri Kalyana Sri Madhuri, Kamireddy Sri Siva Sarojaditya, Mohammad Chisty Madeena Sharieff

Abstract: Cryptocurrency markets are known for their high volatility and complex price dynamics, which make accurate prediction and analysis extremely challenging. Traditional financial forecasting models and classical machine learning algorithms often struggle to capture the nonlinear and rapidly changing patterns present in cryptocurrency datasets. In recent years, advancements in artificial intelligence and quantum computing have opened new possibilities for analyzing complex financial data and improving prediction accuracy.This study proposes a quantum computing–based framework for cryptocurrency market prediction by integrating quantum machine learning techniques with financial time-series analysis. The proposed model utilizes quantum computing concepts such as quantum feature mapping, variational quantum circuits, and quantum recurrent neural networks to analyze cryptocurrency market data. Historical datasets containing information about cryptocurrency prices, trading volume, and market capitalization are used to train and evaluate the model.The proposed system aims to identify hidden patterns in cryptocurrency market trends and generate accurate predictions for future price movements and market volatility. The performance of the quantum-based model is compared with classical deep learning models such as Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks. Experimental results indicate that the quantum machine learning approach achieves improved prediction accuracy and lower forecasting error compared to traditional deep learning models.By leveraging the computational advantages of quantum computing, the proposed framework provides a powerful approach for analyzing highly complex financial datasets. The results demonstrate that quantum machine learning techniques have the potential to significantly enhance cryptocurrency market analysis, enabling more accurate forecasting and better decision-making for investors and financial analysts.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.169

Machine Learning–Based Framework For Accurate CO₂ Emission Prediction And Environmental Impact Analysis

Authors: Mrs.KanakaTulasi P.Reddi, Jittuka Harsha Dinni Sri, Mohan Sai Krishna Bhuvanasi, Adipudi Naga Sai Sri Sowmya, Koruprolu Gowtham

Abstract: The rapid increase in carbon dioxide (CO₂) emissions has become a major environmental concern due to its significant contribution to global warming and climate change. Accurate prediction of CO₂ emissions is essential for developing effective environmental policies and implementing sustainable strategies to reduce greenhouse gas emissions. Traditional statistical forecasting methods often struggle to capture complex relationships between multiple environmental and industrial factors that influence carbon emissions. In recent years, machine learning techniques have emerged as powerful tools for analysing environmental data and improving prediction accuracy.This study presents a machine learning–based framework for forecasting CO₂ emissions using historical environmental and fuel consumption data. The proposed system analyses various factors such as fuel consumption patterns, vehicle characteristics, engine size, and other related attributes to estimate future carbon emissions. Several machine learning regression algorithms, including Linear Regression, Gaussian Process Regression, Multilayer Perceptron (MLP), and Sequential Minimal Optimization for Regression (SMOreg), are implemented and evaluated to determine the most accurate prediction model.The dataset used in this research is obtained from a publicly available environmental dataset and undergoes preprocessing steps such as data cleaning, normalization, and outlier detection to improve model performance. The trained models are evaluated using performance metrics such as Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Relative Absolute Error (RAE), Root Relative Squared Error (RRSE), and correlation coefficient.Experimental results indicate that machine learning algorithms can effectively predict CO₂ emissions, with SMOreg demonstrating superior performance compared to other models in terms of prediction accuracy and error reduction. The proposed framework can assist environmental researchers and policymakers in understanding emission trends and making informed decisions for climate change mitigation.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.170

Explainable Artificial Intelligence For Accurate Household Energy Consumption Forecasting Using Machine Learning Models

Authors: Dr. A.Avinash, Dosapathni Durga Venkata Lakshmi, Rayudu Dona Nikhila, Rayudu Dona Nikhila, Dulla Lokesh Veera Sai Nandan

Abstract: Efficient energy management has become increasingly important due to the growing demand for electricity, rising energy costs, and the need to reduce environmental impact. Accurate prediction of household energy consumption can help individuals and energy providers optimize energy usage, improve resource planning, and promote sustainable living. Traditional statistical forecasting methods often struggle to capture complex consumption patterns present in real-world energy datasets. With the advancement of artificial intelligence, machine learning techniques have shown strong potential for analysing energy consumption data and producing more accurate predictions. This study proposes a machine learning–based framework for predicting household energy consumption using historical electricity usage data. The system analyses various factors such as electrical current, voltage, frequency, and previous energy consumption values to forecast future energy usage. Multiple machine learning and deep learning models, including Linear Regression, Random Forest Regressor, LightGBM, XGBoost, CatBoost, LSTM, and BiLSTM, are implemented and evaluated to identify the most effective model for energy consumption prediction. In addition to prediction accuracy, the proposed framework integrates Explainable Artificial Intelligence (XAI) techniques to improve transparency and interpretability of model predictions. Explainability methods such as Local Interpretable Model-Agnostic Explanations (LIME) and Shapley Additive Explanations (SHAP) are used to analyse the importance of different input features and understand how they influence the prediction results. Experimental results demonstrate that gradient boosting–based models provide highly accurate predictions, while XAI techniques help reveal the key factors that influence energy consumption patterns. The proposed system provides both accurate forecasting and interpretable insights, enabling users to better understand their energy usage behaviour. Such intelligent systems can support energy-efficient decision making, contribute to smart home energy management, and assist in the development of sustainable energy solutions.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.171

Explainable Deep Learning Framework For Brain Tumour Detection And Classification Using MRI Images

Authors: Dr.K.ChandraSekhar, Villa Tejaswi, Vadakattu Lahari Malleswari, Chittavadagi Cristin Pratheek, Mandanakka Surya

Abstract: Brain tumours are one of the most serious neurological disorders that can significantly affect human health and quality of life. Early and accurate detection of brain tumours is essential for effective treatment and improved patient survival rates. Magnetic Resonance Imaging (MRI) is widely used by medical professionals to analyse brain structures and detect abnormalities. However, manual examination of MRI scans can be time-consuming and may lead to inconsistent results due to human interpretation. With recent advancements in artificial intelligence, deep learning techniques have shown great potential in assisting medical experts by automatically analysing medical images.This study presents an intelligent brain tumour detection and classification framework based on deep learning and transfer learning techniques. The proposed system utilizes pre-trained convolutional neural network models to extract meaningful features from MRI images and classify them into multiple tumour categories. Several deep learning architectures, including VGG16, InceptionV3, ResNet50, VGG19, InceptionResNetV2, and Xception, are implemented and evaluated for performance comparison. To improve classification accuracy, an ensemble learning approach is also explored by combining the predictions of the best-performing models.In addition to improving prediction accuracy, the system integrates Explainable Artificial Intelligence (XAI) techniques to provide visual explanations of the regions in MRI images that contribute to the model's predictions. This helps increase transparency and reliability, which are important for medical applications.Experimental results demonstrate that the ensemble-based deep learning model achieves higher accuracy compared to individual models while providing reliable tumour classification results. The proposed framework can assist healthcare professionals in detecting brain tumours more efficiently and may contribute to faster diagnosis and better treatment planning in clinical environments.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.172

Intelligent Crop Recommendation System Using Machine Learning And Deep Learning For Precision Agriculture

Authors: Dr.M.Radhika Mani, P Srinivasa Rama Harshitha, Vangala Vasudev, Sri Sai Vinay Vanaparthi, Gelam Jaya Shankar Krishna Mohan, Angadi Haribabu

Abstract: Agriculture plays a crucial role in ensuring food security and supporting the global economy. However, selecting the most suitable crop for a particular region remains a major challenge for many farmers due to variations in soil nutrients, climate conditions, and environmental factors. Incorrect crop selection can lead to reduced productivity, inefficient use of resources, and financial losses. With the increasing availability of agricultural data and advances in artificial intelligence, machine learning techniques have emerged as powerful tools for improving agricultural decision-making.This study presents an intelligent crop recommendation system that integrates machine learning and deep learning models to assist farmers in selecting the most suitable crop based on soil and environmental conditions. The proposed system analyses important agricultural parameters such as nitrogen (N), phosphorus (P), potassium (K), rainfall, soil pH, temperature, and humidity. These features are used to train predictive models that can recommend the optimal crop for cultivation.Several machine learning and deep learning algorithms, including Decision Tree, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Naïve Bayes, Artificial Neural Networks (ANN), Deep Neural Networks (DNN), and Temporal Convolutional Networks (TCN), are implemented and evaluated. The models are trained using a publicly available agricultural dataset containing multiple crop types and environmental attributes. Performance evaluation is conducted using metrics such as accuracy, precision, recall, and F1-score to determine the most effective model.Experimental results demonstrate that ensemble and deep learning models achieve high prediction accuracy in recommending suitable crops. The system also includes a user-friendly interface that allows farmers to input soil and environmental parameters and receive crop recommendations in real time.The proposed approach contributes to the development of precision agriculture systems by supporting data-driven farming practices, improving crop productivity, and helping farmers make more informed agricultural decisions.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.173

An Intelligent Machine Learning Framework For Cloud Vulnerability Detection And Threat Prevention

Authors: Mrs.Ch.Sowjanya, Kadari Jagadeeswara Veerraju, Yerra Pallavi Rani, Ganni Sameera, Bobbili Lakshmi, Thumu Jayanth

Abstract: Cloud computing has transformed the way organizations store data, deploy applications, and manage digital infrastructure. Its scalability, flexibility, and cost efficiency have made it an essential technology for modern businesses. However, as cloud environments grow in size and complexity, they also become more vulnerable to various cybersecurity threats. Issues such as misconfigurations, insecure APIs, weak authentication mechanisms, and unauthorized access can expose cloud systems to serious security risks. Traditional security mechanisms such as firewalls and rule-based intrusion detection systems often struggle to detect new or evolving threats in dynamic cloud environments.To address these challenges, this work explores the use of machine learning techniques to improve cloud security by predicting and detecting vulnerabilities in distributed systems. The proposed approach analyses security-related data such as system logs, network traffic patterns, and vulnerability reports to identify abnormal behaviour and potential threats. Multiple machine learning algorithms, including Decision Tree, Support Vector Machine (SVM), Random Forest, K-Nearest Neighbors (KNN), and Isolation Forest, are evaluated to determine their effectiveness in detecting security vulnerabilities.The experimental results indicate that ensemble models, particularly Random Forest, provide higher accuracy and better detection capability compared to other algorithms. Machine learning-based security systems can analyse large volumes of data in real time, identify suspicious patterns, and respond to potential threats more quickly than traditional security approaches.By integrating machine learning into cloud security frameworks, organizations can build more proactive and intelligent defence systems capable of adapting to evolving cyber threats. The proposed approach enhances vulnerability detection, reduces response time to security incidents, and supports the development of more resilient and secure cloud infrastructures.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.174

Rewiring Nigeria’s Energy Future: Blockchain And The Possibility Of Peer‑to‑Peer Electricity Trading

Authors: O.B Ayoko

Abstract: Blockchain technology is reshaping how electricity can be produced, traded, and governed, offering new possibilities for countries grappling with unreliable grids and persistent supply gaps. This paper investigates the emergence of blockchain‑enabled peer‑to‑peer (P2P) energy trading, using Nigeria as a lens to explore how decentralized digital infrastructure could redefine participation in electricity markets. Drawing on parallels with the rapid digitalization of financial services, the study examines how distributed ledger systems can support direct energy exchange between prosumers, shift utilities toward roles as market custodians, and improve system trust through transparent, tamper‑proof transaction records. The analysis evaluates regulatory readiness, technical prerequisites, and socioeconomic impacts within Nigeria’s evolving energy ecosystem, where chronic shortages and grid instability create both urgency and opportunity for alternative market models. The findings highlight the potential for P2P trading to accelerate energy access, stimulate local investment, and catalyse a more resilient, consumer‑centric electricity sector.

DOI:

Motivation In The Digital Classroom – High School Students Experiences With Technology-Enhanced Learning In An Israeli Public School

Authors: Amizur Nachshoni

Abstract: This mixed-methods study examines how technology-enhanced learning (TEL) influences student motivation among 11th and 12th-grade students at Golda Meir High School in Ness Ziona, Israel. The research utilized a convergent parallel design to collect both quantitative survey data (n=43) and qualitative open-ended responses from students engaging with Classoos, Google Classroom, Kahoot, and Padlet. Quantitative results demonstrated strong positive trends, with 88.6% of students agreeing or strongly agreeing that technology increases motivation and 91.4% reporting enhanced interactivity. However, 45.7% acknowledged technology-related distractions. Thematic analysis of qualitative data revealed four primary themes: (1) Increased Engagement Through Interactivity and Choice; (2) Autonomy and Access Support Self-Directed Learning; (3) Collaboration and Social Learning Enhance Connection; and (4) Technical and Pedagogical Barriers as Demotivators. The findings suggest that a strategic blend of interactive, collaborative, and autonomy-supportive technology can significantly enhance student motivation when implemented with attention to pedagogical integration and digital distraction management. This study contributes to the understanding of TEL in Israeli secondary education and provides practical implications for educators seeking to optimize technology integration for motivational benefits.

DOI: https://doi.org/10.5281/zenodo.19706653

Digital Twin For Disaster Evacuation Simulation

Authors: Bhargavi Jangam, Yagnavi Rajula, Nivedhika Poloju, Aravind Kumar Kurakula

Abstract: Planning safe evacuation during disasters is extremely important, yet traditional methods are oftenrigid, expensive, and difficult to update. In this paper, we present a Digital Twin–based Disaster Evacuation Simulation System that creates a virtual version of real-world environments such as buildings. The system uses agent-based simulation implemented in Python along with real-time visualization to model how people move during emergencies like fires, floods, or earthquakes. It helps in understanding how congestion forms and how evacuation routes are used under different conditions. By testing multiple scenarios in a virtual setup, the system makes it easier to identify bottlenecks and improve evacuation strategies. Overall, this approach offers a safer and more cost-effective alternative to physical drills and supports better planning for emergency situations.

DOI: https://doi.org/10.5281/zenodo.19443813

Gesture Based Presentation Controller Using Hand Gestures

Authors: Gyara Monika, M. Suchith Reddy, Macha Sujith, Thakur Akshat Singh

Abstract: The project proposes a Gesture-Based Presentation Controller that allows users to control slides using hand gestures through a webcam, eliminating the need for keyboards or remote controllers. It uses computer vision and hand- tracking techniques to detect real-time hand movements and identify key landmarks such as finger positions and hand orientation. Predefined gestures are recognized by analyzing spatial relationships between fingers and joints, ensuring accurate interpretation of user actions. Recognized gestures are mapped to presentation commands like next/previous slide, slideshow control, and pointer movement by simulating keyboard and mouse inputs. The system includes gesture stabilization mechanisms to improve accuracy and is lightweight, cost-effective, and suitable for classrooms, corporate meetings, and professional presentations.

DOI: https://doi.org/10.5281/zenodo.19443936

Ai-Based Hospital Assistance System Using Indian Sign Language Translation

Authors: Srinithi R T, Tisya Chellapandian, Venisha K, Dr. Sumathi V P, Dr. Sumathi V P

Abstract: This research presents a hospital assistance framework that uses AI to enable smooth communication between patients and reception staff. The system recognizes Indian Sign Language (ISL) in real-time and translates speech and text. This helps guide patients effectively without needing a human interpreter. The framework allows for two-way communication: patients use ISL gestures, which are translated into text or voice for the receptionist. In turn, the receptionist’s responses convert back into ISL animations displayed to the patient. The model uses Convolutional Neural Networks (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) architectures, along with a Connectionist Temporal Classification (CTC) decoder for aligning sequences. The preprocessing pipeline uses MediaPipe and OpenCV to extract hand landmarks and reduce noise. A dataset with healthcare-related gestures, such as “doctor,” “appointment,” “medicine,” and “wait,” trained the model. The system operates fully on software and does not require specialized hardware. This solution offers an efficient and accessible way for guiding patients through hospital services, ensuring inclusivity and improving communication at the reception desk.

DOI: https://doi.org/10.5281/zenodo.19444073

Web-based Travel Planning Platform With Integrated AI Chat Assistant

Authors: Ghanshyam G.Lihankar, Sarvesh D.Tak, Akhilesh M.Bhagat, Dhiraj R.Gedam, S V..Raut, D G..Ingle, R S.Durge

Abstract: This research presents the design and development of an integrated web-based travel planning platform combined with an AI-driven chat assistant to improve the efficiency and convenience of travel planning. The system is developed to provide intelligent recommendations, automate itinerary creation, and assist users in making better travel decisions. The paper describes each phase of development, including requirement analysis, system architecture design, module integration, and implementation. The proposed system enables users to interact with the platform through a conversational AI assistant that understands user preferences, travel interests, and budget constraints. Based on user inputs, the system generates personalized travel plans, suggests suitable destinations, recommends accommodations, and provides relevant travel information.

DOI:

Experimental Insights Into Plant Disease Detection: Parametric, Combinatorial, And Computational Evaluations Of Data Mining And Optimization Approaches

Authors: Swapnil Wagh, Ruchi Sharma, Ankit Temurnikar

Abstract: Plant diseases are also known to place huge burden on food security structure and agriculture to the global world; it is approximated that all plant diseases development costs a giant (an estimated 220 billion/year). To address this, the computer vision -specific and deep learning based automated disease detection systems are expandingly viewed as rather interesting as an option instead of the traditional forms of diagnosing that involve a significant amount of new employees . However, the literature screening is saturated with models that have been alleged to be super high in accuracy with regard to classifications when they are under some form of controlled conditions in the laboratory that must in no way imply any trustworthy depiction that they can be relayed over the situation in the real field. It can be said that such discrepancy in performance can stress the idea that there is a dire necessity to carry out more related and stiffer analysis of existing measures of data mining and optimization. This article has such an experimental alloy of which the plant disease variable models can be detected multi faceted in, which is discussed in detail on three axes parametric axis, combinatorial axis, computational axis. The rate of model performances to the hyperparameter options enshrined in the parametric assessment that may also be the optimizers are called counting. The combinatorial work involves the study of connections pertaining to the utility of various Convolutional Neural Network Convolutional designs, as well as the use of spectacular measures of data augmentation and fold up learning methods. The computational verification provided is a practical test of the feasibility of the model, comparison of statistics on the training time, model complexity, and speed of inference. According to the opinion that our experimental findings indicate, our individual models (as well as our EfficientNet) that come with the highest classification performance of about above 98 percent accuracy would always be the best trade off between accuracy and efficiency whereas ensemble models would adopt a combination of soft voting as the best trade off prerogative. The paper further estimates the radical performance augmentation with the generative data augmentation models against the conventional geometric transformations to apply the models in the truly competitive use. The primary accomplishment of this project is the system, which surpasses those pathetic signs of precision and rests upon the familiarization of scientists and performers with how to create, alter, and put to practical practice the scaleable, resilient, and effective plant disease detection methods used in the enhancement of the designated work in the agricultural forerunners.

DOI: https://doi.org/10.5281/zenodo.19444915

Cyberbullying Detection On Social Media Using Compact BERT MODEL And CNN-LSTM

Authors: Pranjal Mahendra Bhosale, Revati Machindra Wahul

Abstract: Cyber bullying is an increasing issue on all online platforms, especially targeting teenagers and young people. Conventional machine learning algorithms fail to perform well in identifying subtle or context-related abusive language. Recent developments in Natural Language Processing (NLP), specifically the transformer model BERT, have demonstrated immense potential in text classification. However, the computational requirements of the full-sized BERT model make it impractical for real-time applications or mobile-based solutions. Proposed in this research is a fast and light cyberbullying detection system based on compact BERT variants like DistilBERT and TinyBERT,CNN,LSTM. These models preserve the language understanding abilities of the original BERT model but with far fewer parameters and computational costs. The model is then fine-tuned on labeled datasets with content related to cyberbullying, and particular emphasis is placed on handling the class imbalance problem through methods such as Focal Loss. Through this process, the model is able to achieve performance metrics that are comparable to those of the full-sized BERT models.

DOI:

SkillBridge: A Digital Solution For Bridging The Gap Between Skills And Employment

Authors: Ishita Shinde, Tanushri Jadhav, Apurva Ransing, Pritesh Patil, Dr. Mrunal Pathak

Abstract: The SkillBridge app aims to link professionals from a variety of industries with people who wish to acquire practical skills. Traditional learning approaches occasionally fall short of offering real-time mentoring and hands-on experience in today’s quickly changing digital world. SkillBridge fills the need of developing a platform where students can find mentors, access skill-based resources and work together on learning opportunities. The software encourages knowledge sharing, community-driven learning, and skill development across a range of professions. Through the use of technology, SkillBridge assists professionals, students, and lifelong learners in increasing the effectiveness, accessibility, and interactivity of skill learning.

DOI: https://doi.org/10.5281/zenodo.19449850

Gesture Vocalizer

Authors: Piyush Pawar, Ayush Jagdale, Yuvraj More, Gunesh Padmukhe, Prof. Meshram A.G

Abstract: The Gesture Vocalizer is a smart assistive communication system developed to help speech- impaired and physically challenged individuals convey messages using hand gestures. The system employs gesture-detection sensors such as flex sensors or accelerometers to recognize predefined hand movements. These gestures are processed by a microcontroller, which converts them into corresponding voice outputs through a speaker or mobile application. The device enables real-time communication without the need for verbal speech, making it highly useful in daily interactions, hospitals, and emergency situations. Users can customize gesture-to- message mappings, improving flexibility and usability. By combining sensor technology, embedded systems, and voice output, the Gesture Vocalizer enhances independence, accessibility, and social interaction for differently-abled individuals.

DOI:

Rentease Connecting Owners And Tenants With Ease

Authors: Poosarla Durga Bhavani, Shaik Roshan Jameer, Vasamsetti Monika Durga Satya Vani, Shaik Ubaid Ahamad, Purushottapatnapu Ravi Sai Krishna, Mr. A. V. Sudhakar Rao

Abstract: Finding a rental or a tenant is often a difficult task. The market is messy, with listings spread across many sites, often outdated, and communication is slow. “Rent Ease” is an all-in-one digital platform built to fix these problems, providing a single, reliable hub that makes renting simpler for everyone. This platform integrates modern technologies such as React.js for frontend development, FastAPI for backend services, and MongoDB for data storage. Key features include real-time chat communication between users and owners, Google Maps integration for accurate location tracking, and an AI-powered module that automatically generates property descriptions to enhance listing quality. Additionally, the system provides filtering options, detailed property views, and a rating and review mechanism to ensure transparency and better decision- making. This system improves user experience, reduces communication gaps, and provides reliable property information, making it a comprehensive solution for both property owners and tenants. Our main goal is to modernize the rental experience. By leveraging technology to connect owners and tenants directly, Rent Ease makes the process faster, more transparent, and significantly less of a hassle, creating.

DOI: https://doi.org/10.5281/zenodo.19451659

An Intelligent Credit Risk Prediction Framework Using Machine Learning Algorithms

Authors: Ms.G.Naga Rani, Mangipudi V N S Sekhar Sarma, Khandavalli V V Lakshmi Srirama Karthik, Malla Karthik, Pabbineedi Vanshika, Ventru Hemanth Kumar

Abstract: The banking sector plays a vital role in the global financial system by providing loans to individuals and businesses for various purposes. While loans generate significant revenue through interest, there is always a risk that borrowers may fail to repay the loan, resulting in financial losses for lending institutions. Therefore, accurately predicting the risk level associated with a loan application is an important task for banks and financial organizations. Traditional loan approval processes rely heavily on manual analysis of customer information, which can be time-consuming and prone to human bias. With the advancement of machine learning techniques, automated systems can now analyse large amounts of financial data to support more efficient and accurate loan approval decisions. This study proposes a machine learning-based loan risk prediction system that analyses customer personal and financial attributes to determine the likelihood of loan default. The dataset used for this study contains multiple features commonly included in loan applications, such as credit history, checking account status, loan amount, employment status, and age of the applicant. Data preprocessing techniques including outlier removal, categorical encoding, and feature scaling are applied to prepare the dataset for model training. Several machine learning algorithms are implemented and compared, including Decision Tree, Random Forest, Support Vector Machine (SVM), Multi-Layer Perceptron (MLP), Naive Bayes, and a Stacking Ensemble model. The models are evaluated using performance metrics such as accuracy, precision, recall, F1-score, and confusion matrix analysis. Experimental results demonstrate that ensemble-based approaches provide improved predictive performance compared to individual machine learning models. The proposed system can assist financial institutions in making faster and more reliable loan approval decisions by identifying high-risk applicants before granting loans. By leveraging machine learning techniques, the system enhances the efficiency of credit risk assessment and supports more effective financial decision-making in the banking industry.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.146

TruthShield-ML – An Intelligent Machine Learning Framework For Accurate Fake News Detection And Misinformation Analysis

Authors: Mrs.K.Ganga Devi Bhavani, Bonam Geetha Chitti Jyothi, Siravapu Santhi Kumari, Karri Manikanta Sai, Alluri Sri Akshay Satya Srinivas, Thella Aditya

Abstract: The spread of fake news has become a significant concern in today’s society, as misleading information can easily damage reputations and lives. To address this issue, researchers have developed fake news detection systems using machine learning techniques. The identification of fake news is rapidly gaining traction and is increasingly being adopted by various industries, either for their own use or to offer as a service to others. Machine learning (ML) and deep learning (DL) are two prominent approaches employed to determine the authenticity of news. There are various methods available for detecting false news through both ML and DL techniques. This paper presents a comprehensive analysis of fake news detection using machine learning approaches. Upon thorough examination, it was found that several ML and DL algorithms have been applied in this domain, with the Support Vector Machine (SVM) being the most commonly used ML method, and Long Short-Term Memory (LSTM) being the most widely applied DL technique.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.147

Deep Learning-Based Intelligent Traffic Violation Detection System Using YOLOv7

Authors: Mr.V.Prem Kumar, Manyam Teja Siva Ganesh Goud, Kothapalli Vennela Sri Sai Bhargavi, Yeluri V N S S P Teja, Vedagiri Yuva Sai Suresh, Dara Teja

Abstract: Traffic violations have become a major cause of road accidents and fatalities in many countries, particularly in densely populated urban areas. Common violations such as red-light jumping, triple riding on two-wheelers, and reckless driving significantly increase the risk of road accidents. Traditional traffic monitoring systems rely heavily on manual observation by traffic police or limited sensor-based systems, which are inefficient, time-consuming, and prone to human errors. To address these challenges, intelligent traffic monitoring solutions based on computer vision and deep learning have gained significant attention. This paper proposes a deep learning-based automated traffic violation detection system using the YOLOv7 object detection model. The proposed system processes video streams obtained from roadside surveillance cameras and analyses them frame-by-frame to detect different traffic violations. The YOLOv7 model is employed to identify vehicles and generate bounding boxes around detected objects. A predefined threshold line is used to determine whether a vehicle crosses the traffic signal during a red light, thereby identifying signal violations. Additionally, the system detects over boarding or triple riding on two-wheelers by analysing the number of riders detected within a single vehicle bounding box. The system uses publicly available datasets such as the MS COCO dataset for vehicle detection and a custom annotated dataset for over boarding detection. The model is trained and evaluated using performance metrics including precision, recall, F-measure, and mean Average Precision (mAP). Experimental results demonstrate that the proposed model effectively detects multiple traffic violations with high accuracy while maintaining efficient real-time performance. The proposed approach provides a cost-effective, automated, and scalable traffic monitoring solution that can assist traffic authorities in improving road safety and reducing the workload associated with manual monitoring systems. The system can be integrated with existing smart city surveillance infrastructures to enhance intelligent transportation management and law enforcement.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.148

An AI-Driven Machine Learning Framework For Accurate Global Solar Radiation Prediction Using Satellite Imagery

Authors: Dr. K. Mounika, Rachakonda Shashikanth, Palivela Prem Chandu, Virothula Sasi Kumar, Balusu Eswar, Kandikonda Pranay

Abstract: The accurate forecasting of Daily Global Solar Radiation (DGSR) is a vital tool in renewable energy planning, climate research, and environmental monitoring. This paper presents a proposal for utilizing machine learning to enhance the estimation of DGSR by using satellite image data. This method utilizes reflectance values obtained from Metaset Second Generation (MSG) satellite images across multiple spectral channels and relies on ground-based meteorological parameters rather than traditional models. Artificial Neural Network (ANN) and Support Vector Machine (SVM) are two supervised machine learning regression models that are utilized for forecasting solar radiation. How do they compare and contrast? The Gharda radiometric station in Algeria has collected measurable solar radiation data for four years (2014-2017) using inputs from satellite imagery, which are then combined to create the dataset. To evaluate these models, statistical performance metrics such as Root Mean Square Error (RMSE), Normalized RMSEA (NRMSE) and MAE (Made Absolute Percentage Error), MBE (Merck-McGregor), and correlation coefficient (R) are utilized. Prediction accuracy is significantly influenced by the number and combination of satellite input parameters, as demonstrated by experimental data. Compared to the SVM, the ANN model had a better RMSE of 21221. The NRMSE, MAPE, and MBE have all been reported with 3.46%, 2.85%, 7.26, etc, respectively. Wh/m2. A 0.99 correlation coefficient is associated with a Wh/m2 value.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.149

Hybrid Physics-Guided Deep Transfer Learning For Accurate Traffic State Estimation

Authors: Dr. Y V Ram Kumar, Keerthi Teja Sri, Patamsetti Yaga Sri Satya Prasanna, Chintapalli Sashank, Theripogu John, Surapureddy Venkata Sai Praneeth

Abstract: Accurately estimating traffic states is a crucial aspect of transportation engineering, enabling effective traffic control and operations. In recent years, Physics-Regulated Deep Learning (PRDL) has gained significant attention due to its ability to achieve higher accuracy while requiring less training data compared to conventional deep learning (DL) approaches. However, a key challenge of PRDL is the lengthy training time required for closely related but distinct tasks.To address this limitation, this paper introduces a hybrid physics-regulated deep transfer learning approach that leverages the strengths of transfer learning, PRDL, and DL to enhance estimation accuracy and reduce computational costs, particularly in scenarios with limited observation data. The proposed framework includes two transfer learning variants designed to extract and transfer essential features from pre-trained models to new but similar traffic environments. This hybrid approach integrates deep learning training, minimizing computational overhead by eliminating physics-based loss calculations during training.Simulation results demonstrate that, compared to traditional PRDL methods, the proposed transfer learning approaches improve estimation accuracy by over 12% on average while reducing training time by more than 50% on average. These findings highlight the potential of hybrid transfer learning techniques in accelerating the adoption of PRDL for traffic state estimation, making it a valuable tool for transportation systems with limited computational resources.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.150

An Explainable Deep Learning Approach For Identification And Classification Of AI-Generated Synthetic Images

Authors: Mrs.M.Uma Devi, Togaru Reshma Sri, Katikidala Satya Ratna Naveen, Gubbala Leela Madhavi, Kalvakolanu Venkata Pavan Chaitanya, Velduti Srivenkata Surya Sai Kumar

Abstract: The rapid advancement of generative artificial intelligence has made it increasingly difficult to distinguish between real images and AI-generated synthetic images. Modern diffusion models can produce highly realistic visuals that closely resemble authentic photographs, raising serious concerns about misinformation, digital fraud, and media manipulation. As synthetic image generation becomes more accessible, reliable detection mechanisms are essential to maintain digital trust and security.This project presents an image classification framework for identifying AI-generated synthetic images using deep learning techniques. A balanced dataset is constructed by combining real images from the CIFAR-10 dataset with synthetic images generated using Stable Diffusion. A Convolutional Neural Network (CNN) model is trained to perform binary classification, distinguishing between real and fake images. In addition to classification, Explainable Artificial Intelligence (XAI) techniques such as Grad-CAM are applied to interpret model decisions and visualize the regions that influence predictions.Experimental results demonstrate that the proposed model achieves high accuracy in detecting synthetic images while maintaining reliable generalization performance. The explainability component further enhances transparency by revealing distinctive patterns and artifacts present in AI-generated images. The proposed system contributes to improving digital image forensics and strengthening defences against AI-driven visual misinformation.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.151

SeaGuard-AI: An Adversarial Robust Framework For Reliable Sea State Estimation In Autonomous Marine Vessels

Authors: Mrs.KanakaTulasi P.Reddi, Sai Varshitha Kuppili, Gabu Ganesh Sasikanth, Adapa Sai Teja Venkata Vinay, Medaboyina Karthik, Trivinesh Gundra

Abstract: Autonomous marine vessels rely heavily on artificial intelligence systems for accurate sea state estimation, which plays a crucial role in navigation, stability control, and operational safety. However, AI-based models are vulnerable to adversarial attacks, where small and carefully crafted perturbations in input data can significantly degrade model performance. Such attacks may compromise safety and reliability, especially in critical maritime environments.This project proposes a novel robustness-enhancing adversarial defence approach to improve the reliability of AI-powered sea state estimation systems. The framework focuses on strengthening deep learning models against adversarial perturbations while maintaining high estimation accuracy. The system integrates adversarial training and defensive mechanisms to enhance model stability under uncertain and hostile conditions. Experimental evaluation demonstrates that the proposed defence strategy significantly improves robustness without sacrificing predictive performance. The results confirm that the enhanced model maintains reliable sea state estimation even in the presence of adversarial disturbances.The proposed approach contributes to improving the safety, security, and reliability of autonomous marine navigation systems.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.152

SecureCPS-Opt: A Hybrid Optimization And Federated AI Framework For Efficient And Privacy-Preserving Attack Detection In IoT-Enabled Cyber-Physical Systems

Authors: Mr.M.Raja Kumar, Pepakayala Bhavani Sri Alekhya, Dasari Asritha, Sada Uma Maheswara Rao, Thimmasatthi Venkateswarlu, Kollu Rajesh

Abstract: The rapid growth of Internet of Things (IoT) devices has significantly improved automation, connectivity, and data-driven decision-making across various domains such as healthcare, smart cities, agriculture, and industrial systems. However, the increasing number of interconnected devices has also introduced serious security challenges. IoT-enabled cyber-physical systems are highly vulnerable to cyber-attacks such as Distributed Denial of Service (DDoS), data injection, botnet attacks, and unauthorized access. Traditional machine learning techniques often struggle to provide high detection accuracy due to imbalanced datasets, high-dimensional features, and inefficient parameter tuning. In this project, a hybrid deep learning-based intrusion detection framework is proposed for identifying security attacks in IoT-enabled cyber-physical systems. The proposed model combines Convolutional Neural Network (CNN) and Deep Belief Network (DBN) to improve feature learning and classification performance. To enhance the model’s efficiency and convergence speed, a novel hybrid optimization technique called Seagull Adopted Elephant Herding Optimization (SAEHO) is employed for tuning the classifier weights. The proposed framework is evaluated using standard IoT intrusion detection datasets such as UNSW-NB15 and BoT-IoT. Performance is measured using metrics including accuracy, precision, sensitivity, specificity, False Positive Rate (FPR), False Negative Rate (FNR), and Matthews Correlation Coefficient (MCC). Experimental results demonstrate that the hybrid classifier optimized using SAEHO outperforms conventional machine learning and optimization-based models in terms of detection accuracy and reduced error rates. The developed system provides an effective and scalable solution for enhancing security in IoT-enabled cyber-physical environments.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.153

BrakeGuard-XAI – An Advanced Secure Explainable AI Paradigm For Early-Stage Brake Anomaly Detection And Interpretable Predictive Maintenance

Authors: Dr. Y. Jayababu, Gokeda Veera Satya Sri Pravallika, Ayyarapu Teja, Appasani Hari Kailash Chowdary, Addanki Yuva Sai Surya Prakash, Garapati Poorna Venkata Ranjit Kumar

Abstract: The study suggests an accessible and secure machine learning model for forecasting brake failures in large commercial vehicles. We support this proposal with evidence. Heavy transport vehicles’ Air Pressure System (APS) is constantly monitored by IoT-based sensors in modern day heavy transport systems, generating vast amounts of operational data. Detecting brake failures manually with large and highly unbalanced datasets is time-consuming and inefficient. Our approach to these problems involves the use of K-Nearest Neighbour (KNN) imputation for missing values and SMOTE for dealing with class imbalance. Both methods are effective in both situations. Logistic Regression, Decision Tree, Support Vector Machine, Gradient Boosting, and Random Forest are among the machine learning algorithms that undergo stratified cross-validation during implementation and evaluation. The Random Forest classifier’s accuracy, precision, recall, F1-score and ROC-AUC are shown to be more than satisfactory using experimental data. Enhanced transparency and trust in the prediction process are achieved through the use of Explainable Artificial Intelligence (XAI) techniques like SHAP and LIME, which can interpret model decisions. They also use methods of selecting features that reduce computational complexity while preserving high levels of accuracy in making predictions. This proposed framework improves fault detection reliability, reduces maintenance costs and allows for predictive maintenance in heavy transport systems.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.154

A Deep Learning–Based Multi-Layer Recursive Neural Network Framework For Intelligent Thyroid Disease Detection And Recognition

Authors: Dr. A.Avinash, Kanchi Dhanusha, Saladi Rudra Naga Prasanna Lakshmi, Thiguti Sri Ajitesh, Kesanakurthi Satya Karthikeeyan, Vangapandu Lokesh

Abstract: Thyroid disease is one of the most common endocrine disorders affecting millions of people worldwide. The thyroid gland plays a crucial role in regulating metabolism, growth, and overall body functions. Any imbalance in thyroid hormone production can lead to conditions such as hypothyroidism or hyperthyroidism. Early detection of thyroid disorders is important to prevent serious health complications and to ensure timely treatment. Traditional methods of diagnosing thyroid disease rely on laboratory tests and manual evaluation, which may be time-consuming and sometimes prone to errors. With the advancement of artificial intelligence, deep learning techniques can assist medical professionals in improving diagnostic accuracy and reducing workload. In this project, a deep learning-based Multi-Layer Recursive Neural Network (ML-RNN) is proposed for thyroid disease detection and classification. The system includes data preprocessing, feature selection using the Fisher Score method, and classification using the ML-RNN model. The dataset used for analysis is obtained from a standard repository and includes various thyroid-related attributes. The performance of the proposed model is evaluated using metrics such as accuracy, recall, precision, and error rate. Experimental results show that the ML-RNN model achieves better performance compared to traditional machine learning algorithms such as Support Vector Machine (SVM), K-Nearest Neighbours (KNN), and Random Forest (RF). The proposed approach provides an effective and reliable method for thyroid disease detection.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.155

HEAL (Heatmap For Environmental Air Levels)

Authors: Dheeraj Patil, Sanika Dixit, Aditya Dixit, Meenakshi Deotare

Abstract: Air pollution is one of the most serious environmental threats in urban areas, affecting both human health and climate. Traditional air quality monitoring systems provide only point-based information; hence, this limits their ability to show distributions across a city. Herein, this work describes HEAL, a web-based system for pollution hotspot predictions and visualizations through the utilization of machine learning and data visualization techniques. This system collects air quality data from APIs or sensors, processes it, and generates dynamic heat maps that showcase the levels of pollution in real time. Interpreting the interaction among environmental, traffic, and meteorological data, HEAL offers citizens, policymakers, and researchers new localized insights into air quality variations, which will result in better decision-making.

DOI:

Pothole Detection And Automated Reporting System Using Computer Vision

Authors: Sparsh S. Misal, Yash R. Lodha, Parth P. Gargote, Shivam S. Daundkar

Abstract: Road infrastructure plays a critical role in transportation, but issues like potholes significantly affect safety, efficiency, and maintenance costs. Traditional pothole detection methods rely heavily on manual inspection and public reporting, which are often delayed and inefficient. This project proposes a smart pothole detection and reporting system using computer vision and machine learning. The system uses a live webcam feed to detect potholes in real-time using a model trained with Teachable Machine and deployed using TensorFlow.js. When a pothole is detected, the system captures an image, records the location, date, and time, and automatically generates a complaint ticket. The backend, built using Flask, stores the report data and provides a history of detected potholes. This system offers a low-cost, scalable, and automated solution that can be extended for smart city applications and real-time road monitoring systems.

MentorAI: A Smart Web-Based Learning Assistant With Personalized Guidance And Interactive Study Support

Authors: Dadarkar Ehaan Mubasshir, Ansari Ayan Atif Masood Iqbal, Khan Zain MD Irfan

Abstract: MentorAI is a comprehensive, web-based intelligent learning assistant designed to transform how students study, retain knowledge, and engage with educational content. The platform integrates a personalized AI Tutor powered by large language models, a Voice Recall system for active retrieval practice using the Web Speech API, an SM-2 algorithm-driven Spaced Repetition Flashcard engine, an AI-generated adaptive Quiz Engine, an AI- assisted rich-text Workspace for notes and PDF imports, a 3D Knowledge Graph for visual concept mapping using D3.js force-directed visualization, and a command-palette-style Nexus navigation system. A central Dashboard aggregates learning metrics, study streaks, mastery scores, and AI-detected weak topics in real time. This paper presents the complete system architecture, feature design rationale, technology stack, database design, security model, testing methodology, and results achieved during the development of MentorAI as a final year engineering capstone project.

Quantum-Driven Vector Fusion Networks For Early Cancer Detection Using Machine Learning

Authors: Mr.S.K. Sankar, Janga Sanjay, Didde Vaishali, Dwarampudi Tejo Madhuri, Chitturi Nikhitha

Abstract: Early detection of cancer plays a crucial role in improving patient survival rates and enabling effective treatment strategies. However, traditional diagnostic methods often face challenges such as high-dimensional biomedical data, feature redundancy, and computational inefficiency. Recent advancements in machine learning have improved diagnostic capabilities, but conventional algorithms still struggle to efficiently process complex genomic and medical imaging datasets. To address these limitations, this study proposes a novel framework that integrates quantum computing with machine learning techniques for enhanced cancer detection. The proposed approach employs a sequence of intelligent modules including Quantum-Normalized Adaptive Refinement (Q-NAR) for data preprocessing, Wrapper Component Attribute Analysis (WCAA) for feature ranking, and Swing L-Bee Mustard Optimization (SLBMO) for selecting the most relevant features. Finally, a hybrid predictive model known as the Quantum Boosted Vector Fusion Network (QBVFN) is utilized to perform cancer prediction and treatment outcome analysis. The framework is evaluated using the Cancer Genome Atlas (TCGA) dataset in a Python environment. Experimental results demonstrate significant improvements in feature optimization, prediction accuracy, and computational efficiency for early-stage cancer detection. This research highlights the potential of quantum-assisted machine learning techniques to support next-generation intelligent cancer diagnostic systems.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.139

Ensemble Machine Learning Approach For Urban Flood Hazard Assessment And Risk Mapping

Authors: Mrs.T.N.V. Durga, Kona Lasya, Golla Vidya Prasanthi, Allam Hema Siva Sankar, Kola Amrutha Lakshmi

Abstract: Flooding is one of the most destructive natural hazards, particularly in urban environments where population density and infrastructure development increase vulnerability to extreme weather events. Accurate identification of flood-prone areas is essential for effective disaster management and urban planning. This study presents an ensemble machine learning framework for urban flood hazard assessment by integrating multiple predictive models. The proposed approach combines the strengths of individual machine learning algorithms such as Classification and Regression Trees (CART), Random Forest (RF), and Boosted Regression Trees (BRT) to generate a more reliable flood susceptibility map. Several environmental and geographical factors, including slope, elevation, rainfall, land use, and distance to rivers, are analysed to evaluate their influence on flood occurrence. The ensemble model aggregates the predictions of individual models using weighted averaging techniques to improve prediction accuracy and reduce model bias. Experimental results demonstrate that the ensemble approach outperforms individual models in terms of predictive performance and reliability. The generated flood hazard maps provide valuable insights for identifying high-risk zones and supporting decision-makers in developing effective flood mitigation strategies.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.140

Hybrid Deep Learning Framework for Android Malware Detection Using Application Permissions and Social Media Threat Intelligence

Authors: Mrs.M. Saranya, Parimi Sai Neeraja, Akumarti Venkat, Sistu Sai Purna Sriram, Makada Ravikiran, Padala Chaitanya

Abstract: The rapid growth of smartphone usage has made mobile devices an essential part of everyday life, supporting activities such as communication, online banking, education, and social networking. However, the increasing popularity of Android-based devices has also made them a major target for cyber attackers who develop malicious applications to exploit system vulnerabilities and steal sensitive information. To address this challenge, an intelligent malware detection and prevention framework for Android devices is proposed. The proposed system integrates real-time threat intelligence gathered from social media platforms with deep learning-based malware classification techniques. Malware signatures shared through social media sources are periodically collected and stored in a centralized malware hash database to ensure the system remains updated with newly discovered threats. In addition, the system employs a deep learning model based on a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) architecture to analyze Android application permissions and classify applications as benign or malicious. By combining real-time malware signature updates with deep learning-based behavioural analysis, the proposed framework enhances the accuracy and efficiency of Android malware detection. Experimental evaluation demonstrates that the system achieves high detection accuracy and provides a robust solution for protecting Android devices against emerging malware threats.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.141

Artificial Intelligence-Based Framework For Automatic Detection Of Dysarthria Severity Levels Using Speech Analysis

Authors: Ms.MD.Apsar Jaha, Gollapalli Mounika Subhash Chandra, Govindaraju Sri Lakshmi Swathi, Soorneedi Navaneeth Preetham, Nagulla Seshu Pavan, Kanumenu Siva Senkara Varaprasad

Abstract: Speech disorders significantly affect an individual’s ability to communicate effectively and reduce overall quality of life. Dysarthria is a neurological speech disorder that results from damage to the nervous system and affects the muscles involved in speech production. Traditional assessment of dysarthria severity is usually performed by speech-language pathologists through perceptual evaluation, which can be subjective and time-consuming. Recent advancements in artificial intelligence and machine learning have enabled the development of automated systems capable of analysing speech characteristics and identifying different levels of dysarthria severity. This study presents an overview of intelligent techniques used for the automatic detection and classification of dysarthria severity levels. The proposed approach focuses on analysing speech features such as acoustic patterns, prosodic characteristics, and spectral features extracted from speech signals. Machine learning and deep learning models are then used to classify the severity of dysarthria based on these extracted features. By utilizing AI-based models, the system can provide objective and efficient evaluation of speech impairments. The proposed framework can assist clinicians in improving diagnostic accuracy and developing personalized rehabilitation strategies for individuals affected by dysarthria.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.142

Machine Learning-Based Cyber Attack Detection Framework For Secure Unmanned Aerial Vehicle (UAV) Communication Networks

Authors: Dr Manjula Devarakonda Venkata, Vasa Neeharikasri, Vudatha Rama Subrahmanyam, Suravarapu Venkatesh, Malagala Pavan, Mattaparthi Jaya Praneeth

Abstract: Unmanned Aerial Vehicles (UAVs), commonly known as drones, are increasingly used in various applications such as surveillance, logistics, environmental monitoring, and disaster management. Despite their numerous benefits, the rapid adoption of UAV systems has introduced significant cybersecurity challenges. UAV communication networks are vulnerable to different types of cyber threats including GPS spoofing, data injection attacks, and network intrusions, which can compromise system functionality, mission objectives, and data security. To address these challenges, this study proposes a machine learning-based framework for detecting cyber attacks in UAV systems. The proposed approach combines supervised and unsupervised learning techniques to analyse UAV telemetry data, communication signals, and operational parameters in real time. By performing behavioural analysis and anomaly detection, the system can identify abnormal patterns and isolate potential cyber threats with high accuracy and minimal false positives. Experimental evaluation demonstrates that the proposed framework can effectively detect various attack scenarios while maintaining efficient response time and reliable performance. The integration of machine learning techniques into UAV cybersecurity systems provides a robust solution for enhancing the safety and reliability of drone communication networks.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.143

Regional Wind Power Forecasting Using Bayesian Feature Selection And Machine Learning Techniques

Authors: Mr.Y.Manas Kumar, Sathi Chaitanya Sai Durga, Kollu Ruby Sophia, Gaduthuri Alekhya, Nalluri Lishitha Devi, Pallala Sasi Kiran Reddy

Abstract: The rapid growth of renewable energy sources has increased the importance of accurate wind power forecasting for reliable power system operation. Wind power generation is inherently variable due to changing weather conditions, making prediction a challenging task. This paper presents an intelligent wind power forecasting framework based on Bayesian Feature Selection combined with machine learning models. The proposed approach processes numerical weather prediction data and removes irrelevant spatial features to improve prediction accuracy. A dimensionality reduction technique is applied to select the most informative sub-areas of weather data, thereby reducing computational complexity while maintaining important predictive information. Various machine learning algorithms such as Support Vector Machines, Artificial Neural Networks, and Convolutional Neural Networks are employed for forecasting regional wind power output. The proposed model enhances prediction performance by optimizing feature selection and improving model efficiency. Experimental evaluation demonstrates that the system significantly improves forecasting accuracy while reducing the dimensionality of input data. The framework can assist energy providers and power grid operators in planning and managing renewable energy resources more effectively.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.144

Analysis And Classification Of Adversarial Machine Learning Attacks Against Machine Learning-Based Network Intrusion Detection Systems

Authors: Mr.Y.H.S.S. Phaneedra, Polisetty Nikhitha Sowmya, Kolla Triveni, Garaga Naveen Kumar, Kadali Nikitha Sri Satya Gayatri

Abstract: Network Intrusion Detection Systems (NIDS) play a critical role in modern cybersecurity infrastructures by monitoring network traffic and identifying suspicious or malicious activities. In recent years, machine learning techniques have significantly improved the performance of intrusion detection systems by enabling automated traffic analysis and anomaly detection. However, the integration of machine learning into security systems also introduces new vulnerabilities that can be exploited by attackers. One such threat is adversarial machine learning, where malicious actors manipulate training or testing data to deceive machine learning models and degrade their performance. This study presents a comprehensive analysis of adversarial machine learning attacks targeting network intrusion detection systems. The work explores how adversarial samples are generated by introducing small perturbations into original datasets, which results in incorrect predictions by the intrusion detection model. Furthermore, the paper classifies adversarial attacks based on several criteria, including attacker knowledge level, misclassification objectives, affected learning phase, and the intended security violation. Understanding these attack strategies is essential for designing more robust and secure intrusion detection systems capable of defending against adversarial manipulation.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.145

Sustainable Highway Pavement Design Using Waste Tyre Reinforced Subgrade Materials

Pipe Inspection Robot

Authors: Mahbub Alam, Rakesh Kumar Paswan, Nitish Ravidas, MD Sohail Ahmad, Mr. S.B. Patil

Abstract: Pipeinspectionrobotsare designedto monitorandassess the conditionof pipelinesin industriessuch as oil and gas, watersupply,andsewage systems. Theserobotscan travel inside pipes to detectcracks,blockages,corrosion,andotherstructuraldefectsthatare difficult or dangerousfor humansto inspect.Equippedwith sensors,cameras,andwireless communicationsystems,the robotcapturesreal-timedata and images of thepipeline’ s internalcondition.This technologyimprovesinspectionaccuracy,reducesmaintenancecosts, and enhancessafety by minimizingthe needfor manualinspection in confinedspaces.

IoT Based Crop Monitoring System

Authors: Manoj, Anmol Dobriyal, Suneet Bhalla, Utsav

Abstract: Agriculture in India faces significant challenges due to climate variability, inefficient resource utilization, and lack of real-time field monitoring. Traditional methods rely on manual observation, which often results in delayed detection of critical environmental changes affecting crop health. This paper presents an Internet of Things (IoT) based crop monitoring system designed to provide continuous real-time monitoring of key agricultural parameters. The system integrates soil moisture, temperature, humidity, and gas sensors with an Arduino Nano and NodeMCU for data acquisition and processing. A Global System for Mobile Communications (GSM) SIM800A module is utilized to transmit alert messages when any parameter exceeds predefined threshold values, ensuring remote accessibility without dependence on internet connectivity. A 16×2 Liquid Crystal Display (LCD) provides local visualization of sensor data. The proposed system focuses on simplicity, low cost, and reliability, making it suitable for deployment in rural areas with limited infrastructure. Experimental observations demonstrate that the system accurately monitors environmental conditions and generates timely alerts, enabling improved decision-making and reducing potential crop losses. The solution provides a practical approach toward enhancing agricultural monitoring using IoT technologies in resource-constrained environments.

DOI:

Smart Blood Donor Finder System

Authors: Mr.M.Thangadurai, Lathishna R, Jamuna N, Madhunisha S

Abstract: The Smart Blood Donor Finder System is an efficient and technology-driven solution designed to connect blood donors with recipients in real time. The system aims to address the critical challenge of blood shortages by creating a centralized digital platform where donors can register their details, including blood group, location, and availability. When a request is made, the system quickly identifies suitable donors based on compatibility and proximity, ensuring faster response during emergencies. It utilizes database management, location-based services, and communication technologies such as SMS or notifications to alert potential donors instantly. The system also maintains donor history, eligibility status, and previous donation records to ensure safety and reliability. By reducing manual effort and delays in searching for donors, this system improves efficiency in healthcare services. Overall, the Smart Blood Donor Finder System enhances accessibility, saves time, and increases the chances of timely blood availability, ultimately contributing to saving lives.

DOI: https://doi.org/10.5281/zenodo.19470987

Design And Development Of A Smart Blood Centrifugation System Using AI- Based Process Monitoring And IoT Alerts

Authors: Dr. J. Yogapriya, Johithasri A K, Keerthika A

Abstract: Blood centrifugation is a crucial step in diagnostic testing; however, many rural and low-resource healthcare centers lack access to standard centrifuge machines due to high cost and maintenance requirements. This project proposes an ultra-affordable, portable smart blood centrifuge integrated with Artificial Intelligence and Internet of Things technologies. The device employs a high-speed motor to separate blood components, while an embedded camera continuously monitors the separation process. AI algorithms analyze the captured images to accurately determine the completion of plasma separation, preventing over- or under-centrifugation. An IoT module enables real-time monitoring and status notifications through a mobile application, allowing safe and remote operation. By reducing dependence on skilled manpower and conventional laboratory infrastructure, this intelligent system enhances diagnostic reliability and supports timely clinical decision-making, making it highly suitable for primary healthcare centers, mobile medical units, and underserved regions.

DOI: https://doi.org/10.5281/zenodo.19471072

Early Brain Disease Detection Using Deep Learning And Medical Imaging

Authors: Mrs.L.Nivetha, M.Tharunsuriya, R.Sharugas, S.Vaitheesh

Abstract: The primary objective of this proposed research is to develop a new deep learning algorithm that can analyze neuroimaging data for early detection and diagnosis of brain diseases such as epilepsy, Parkinson’s disease, Alzheimer’s disease, and brain tumors. The algorithm will be developed using a combination of supervised and unsupervised learning techniques. The dataset will include a large number of neuroimaging scans, including MRI, CT, and PET scans, from patients with different brain diseases as well as healthy controls. The algorithm will be trained to differentiate between healthy and diseased brain scans and to classify different types of brain diseases based on the patterns observed in the neuroimaging data. The proposed algorithm will incorporate advanced deep learning techniques such as convolutional neural networks (CNNs), recurrent neural networks (RNNs), and autoencoders, which are specifically designed for processing large and complex datasets. The algorithm will also use transfer learning, which involves transferring knowledge learned from one task to another, to enhance the accuracy of the classification model. The proposed algorithm will be able to detect subtle changes in brain structure and function that may not be visible to the naked eye, enabling earlier detection and diagnosis of brain diseases. The proposed algorithm has the potential to significantly improve the accuracy and speed of diagnosis of brain diseases, leading to earlier and more effective treatment. It could also help identify new biomarkers for brain diseases, leading to a better understanding of the underlying mechanisms and potential new targets for therapy. Ultimately, the proposed algorithm could improve the quality of life for millions of people around the world who suffer from brain diseases such as epilepsy, Parkinson’s disease, Alzheimer’s disease, and brain tumors.

DOI: https://doi.org/10.5281/zenodo.19471133

Reliable Navigation In GPS-Denied Environments Using Doppler Assistance

Authors: Dr. J. Yogapriya, Sangsai ST, Praveen R, Naresh P

Abstract: Reliable navigation in environments where Global Positioning System (GPS) signals are unavailable or degraded remains a critical challenge for autonomous systems, defense operations, and underground or indoor applications. This research proposes a robust navigation framework that leverages Doppler-based velocity estimation to enhance positioning accuracy in GPS-denied environments. The system integrates inertial measurement units (IMUs) with Doppler shift observations derived from radio frequency or acoustic signals to provide continuous and drift-reduced localization. A sensor fusion approach, combining Extended Kalman Filtering and machine learning-based error correction, is employed to mitigate accumulated drift and measurement noise. The proposed model is evaluated in complex scenarios such as urban canyons, tunnels, and indoor settings, demonstrating improved trajectory estimation and resilience compared to conventional inertial-only methods. Experimental results indicate that Doppler-assisted navigation significantly enhances reliability, reduces positional error, and ensures continuous operation in challenging conditions. This approach offers a scalable and efficient solution for next-generation navigation systems in autonomous vehicles and robotics.

DOI: https://doi.org/10.5281/zenodo.19471216

Vision Based Navigation Assistant Using Object Detection And Depth Estimation

Authors: Mrs.S.Subha, Kiruthika M, Harrshinee L, Kanika V

Abstract: Vision-based navigation has become increasingly important in fields such as assistive technology, robotics, and autonomous driving, as it enables systems to understand and interact with complex environments. This study introduces a Vision-Based Navigation Assistant that combines object detection with depth estimation to improve real-time awareness and navigation safety. The system utilizes deep learning techniques to detect and categorize surrounding objects while estimating their distances through monocular or stereo vision approaches. This integrated method allows the system to deliver relevant information about obstacles, pathways, and potential risks. Designed for efficiency, the framework can run on embedded devices, ensuring portability and minimal processing delay. Furthermore, it provides feedback through audio or haptic signals, making it especially useful for visually impaired individuals and autonomous systems. Experimental evaluations indicate enhanced accuracy in identifying objects and estimating distances, resulting in dependable performance across both indoor and outdoor settings. Overall, the proposed solution demonstrates the effectiveness of computer vision in developing intelligent navigation aids that enhance mobility, safety, and user independence.

DOI: https://doi.org/10.5281/zenodo.19471270

ResQHer – A LoRa-Based Smart Womens Safety System

Authors: Dr.S.Dhanabal,M.E.,Ph.D., Amittha K, Deeksha T, Deepshika M

Abstract: Women’s safety has become a significant social and technological concern, particularly in remote and rural regions where immediate communication during emergencies is often unreliable due to poor network connectivity. Most existing safety solutions depend on cellular networks, which may fail in such environments, creating a critical need for an alternative approach. This project proposes ResQHer, a smart women’s safety device that employs LoRa-based hybrid communication to ensure long-range, low-power, and reliable transmission of distress alerts even in areas with limited or no internet access. The system integrates an ESP32 microcontroller, a LoRa SX1278 module for long-distance communication, GPS for real-time location tracking, and GSM/Wi-Fi as backup channels. When the emergency trigger is activated, the device instantly sends alert messages along with precise location details to predefined contacts and a centralized monitoring gateway. Designed to be compact, wearable, and energy-efficient, the device is suitable for everyday use and supports future enhancements such as continuous tracking, data logging, and integration with emergency response services. Overall, ResQHer aims to improve women’s personal safety by enabling faster assistance, dependable communication, and enhanced situational awareness across both urban and rural environments.

DOI: https://doi.org/10.5281/zenodo.19475804

Smart Fault Detection And Recovery System For Industrial Machinery

Authors: Dr.R.Shankar, S.Pooja, K.Sona, S. Harshini

Abstract: Industrial motors and rotating machines play a vital role in manufacturing and production environments, where unexpected failures can lead to significant downtime, financial loss and safety risks. Traditional maintenance approaches such as reactive maintenance, performed after a failure and preventive maintenance, based on fixed time intervals are inefficient and often fail to detect early- stage faults. Existing monitoring systems are expensive, complex and mostly suitable only for large-scale industries, making them inaccessible for small and medium enterprises. Hence a low-cost ESP32-based predictive maintenance system for real-time condition monitoring of industrial motors continuously monitors key health parameters such as vibration, temperature and current using multiple sensors. By analyzing these parameters in real time, the system can detect abnormal operating conditions at an early stage. Fault severity is classified into normal, warning and critical levels, which are indicated using visual LED alerts. In addition, the system provides a self-protection mechanism by automatically disconnecting the motor during severe fault conditions, preventing permanent damage.

DOI: https://doi.org/10.5281/zenodo.19476141

District-Level Crop Yield Prediction Using Government Open Data And AI Techniques

Authors: Ambuj Kumar Misra

Abstract: Accurate crop yield prediction is essential for food security, agricultural planning, and policy formulation. This research paper presents a comprehensive analysis of district-level crop yield prediction using government open data and artificial intelligence techniques [1]. The study leverages publicly available datasets from agricultural ministries, meteorological agencies, and remote sensing sources to develop predictive models utilizing machine learning and deep learning approaches. Our analysis demonstrates that ensemble methods combining multiple algorithms achieve superior accuracy compared to individual models, with R² values exceeding 0.85 on validation datasets [2]. The proposed framework integrates soil characteristics, weather patterns, crop management practices, and historical yield data to create robust prediction systems deployable across different geographical regions [3]. Results indicate that incorporating remote sensing data and temporal patterns significantly improves model performance [4]. This research contributes to the growing body of knowledge on precision agriculture and provides practical guidelines for government agencies and farmers to optimize yield forecasting systems.

DOI: https://doi.org/10.5281/zenodo.19479317

Ai Startup Idea Validator Using Ml And Llm Agents

Authors: Durgunala Ranjith, K.Hari Krishna, K.Rajender, R.Koti

Abstract: The project proposes an AI Startup Idea Validator that helps users evaluate startup ideas automatically using Artificial Intelligence and Large Language Models (LLMs). The system allows users to input their startup ideas through a web interface and analyzes them by considering factors such as market potential, competition, feasibility, and innovation. It uses AI-based processing to generate outputs including feasibility score, SWOT analysis, and improvement suggestions, providing users with clear insights into the strengths and weaknesses of their ideas. The system integrates external data sources and intelligent models to ensure accurate and data-driven decision-making. It is designed to be fast, cost-effective, and user- friendly, making it suitable for students, entrepreneurs, and startup incubators.

DOI: https://doi.org/10.5281/zenodo.19553722

Voiceguard – Ai-Based Voice Authenticity Detection System

Authors: Dr. C. Saravanabhavan, Akhil R

Abstract: Recent advances in deep learning have en-abled highly realistic synthetic speech, creating serious risks such as impersonation, fraud, and misuse of voice-based authentication systems. Detecting AI-generated speech is increasingly difficult because modern text-to-speech and voice conversion models can closely imitate human prosody and timbre across languages. This paper proposes VoiceGuard, a hybrid deep learning framework that combines complementary spectral and temporal rep-resentations for deepfake voice detection. A Convolutional Neural Network (CNN) branch learns frequency-domain artifacts from spectrograms, while a CNN-GRU branch models temporal inconsistencies from acoustic descriptors. An attention-based fusion mechanism adaptively weights branch outputs to improve discriminative power. The framework is evaluated on benchmark datasets and cross-lingual settings, and it improves performance compared to single-representation approaches while remaining compu-tationally practical for real-world deployment.

DOI: https://doi.org/10.5281/zenodo.19480877

Fraud Shield-UPI: The Secure UPI Fraud Detection System

Authors: P. Saranya, Ms. E. Sheela

Abstract: The rapid expansion of digital payment platforms has significantly transformed financial transactions worldwide. In India, the Unified Payments Interface (UPI) has emerged as one of the most widely adopted real-time payment systems due to its speed, convenience, and low transaction cost. However, the increasing popularity of UPI has also led to a substantial rise in fraudulent activities, including phishing attacks, unauthorized fund transfers, identity theft, and account takeover incidents. Traditional rule-based fraud detection systems rely on static thresholds and predefined heuristics, which are often unable to adapt to evolving fraud patterns and complex transaction behaviors. Furthermore, fraud detection datasets are typically highly imbalanced, where fraudulent transactions represent only a small fraction of the total data, making accurate detection more challenging. To address these limitations, this study proposes FraudShield-UPI, a machine learning-based fraud detection framework designed to improve the accuracy and reliability of fraud identification in digital payment systems. The proposed framework integrates Synthetic Minority Oversampling Technique (SMOTE) to handle class imbalance, Principal Component Analysis (PCA) for dimensionality reduction, and Extreme Gradient Boosting (XGBoost) for high-performance classification of fraudulent transactions. The system is implemented as a web- based application using the Flask framework, enabling real-time fraud prediction and interactive transaction analysis. In addition to the proposed model, a comparative evaluation platform is developed to benchmark traditional machine learning algorithms including Decision Tree, Support Vector Machine (SVM), and Random Forest using the same dataset and evaluation metrics. Experimental evaluation on a simulated UPI transaction dataset demonstrates that the proposed SMOTE-PCA-XGBoost model significantly outperforms baseline models in terms of accuracy, precision, recall, and F1-score, while effectively reducing both false positives and false negatives. The results highlight the capability of the proposed framework to detect fraudulent transaction patterns with improved reliability. The modular architecture and web-based deployment further demonstrate the practical feasibility of integrating the system into real- world financial platforms for enhanced digital payment security.

DOI: https://doi.org/10.5281/zenodo.19483516

From Regulatory State To Regulatory Space: Mapping India’s Fragmented Ai Governance Through The Lens Of Comparative Regulatory Theory

Authors: Shailja Jha

Abstract: The rapid proliferation of Artificial Intelligence (AI) technologies has exposed significant limitations in traditional state-centric regulatory frameworks, particularly in complex and diverse jurisdictions such as India. This paper advances the concept of a transition from a “regulatory state” to a “regulatory space,” emphasizing the distributed, multi-actor nature of AI governance. Drawing on comparative regulatory theory, the study analyzes how India’s AI governance is characterized by institutional fragmentation, overlapping mandates, and sector-specific regulatory interventions rather than a unified legal framework. By examining key regulatory bodies, policy instruments, and emerging guidelines across domains such as data protection, digital markets, and sectoral compliance, the paper maps the contours of India’s evolving AI governance ecosystem. It further compares India’s approach with global models, including the European Union’s risk-based regulatory regime and the United States’ market-driven governance structure, to highlight divergences and convergences in regulatory philosophy. The analysis demonstrates that India’s fragmented governance structure, while often viewed as a limitation, may also function as a flexible “regulatory space” that enables adaptive, context-sensitive oversight. However, this flexibility comes with challenges related to coordination, accountability, and enforcement consistency. The paper concludes by proposing a hybrid governance model that integrates centralized policy direction with decentralized regulatory innovation, thereby aligning India’s AI governance with both domestic priorities and global regulatory trends.

DOI: https://doi.org/10.5281/zenodo.19481225

SMARTLOFO – AI Powered Lost And Found Platform

Authors: Mrs. D. Srilatha, Shaik Umaiza Bhanu, Yanamala Lalith, Shaik Amin Sadik, Kamunuri Kasi Ganesh

Abstract: — In the digital era, managing lost and found items efficiently remains a challenge due to reliance on manual methods and unstructured reporting systems. Traditional approaches such as notice boards and text- based communication often result in disorganized data, delayed responses, and low matching accuracy. These limitations highlight the need for an intelligent and automated solution. This paper presents SMARTLOFO: AI Powered Lost and Found Platform, a full- stack web application designed to streamline the process of reporting, tracking, and retrieving lost items. The system is developed using React for the frontend and a Python-based FastAPI backend, with MongoDB/SQLite for data storage. It provides a user-friendly interface along with secure authentication using JWT and bcrypt. A key contribution of the system is the integration of an AI-powered smart matching algorithm. Using Google Gemini, the system performs image analysis to extract item descriptions, categories, and features. These attributes are processed using a scoring-based matching mechanism that evaluates similarity based on category, extracted features, location, and time proximity. Matches exceeding a defined threshold are automatically identified, and users are notified via an email notification system. The platform is deployed on a cloud environment, enabling real-time interaction and accessibility. Despite its advantages, the system depends on user participation and input accuracy. Future enhancements include improving scalability and incorporating advanced machine learning models. Overall, SMARTLOFO demonstrates an intelligent and scalable approach to modernizing lost-and-found systems using artificial intelligence and full-stack technologies.

DOI:

Smart And Intelligent Web Traffic Analytics And Monitoring System

Authors: Mr. Durgunala Ranjith, Gaddam Abhinay Reddy, Guda Raja Krishna, Tejavath Nithin Nayak

Abstract: The “Smart and Intelligent Web Traffic Analytics and Monitoring System” is designed to track and analyze website traffic in a simple and effective way. It collects real-time data about users, page visits, and browsing behavior. The system helps website administrators understand how users interact with their website. It can identify traffic patterns and detect unusual or suspicious activities. Visual reports and dashboards make the data easy to read and interpret. This system supports better decision-making to improve website performance and security.

Predictive Risk Analytics In Project Management Using Graph-Based Lightweight AI And Counterfactual Risk Mitigation

Authors: S. Balaji, N. Poyyamozhi

Abstract: Currently, the field of project management faces increasing uncertainty as projects must deal with changing requirements, resource shortages, and the unpredictable effects of human actions, technical systems, and external events. However, existing data-driven models have failed to provide interpretable results, preventing project managers from identifying the factors that create risks. Thus, this research presents a lightweight and explainable data-driven decision support system that enables project risk prediction and risk management in complex project management environments. The devised methodology employs a Project Management Risk Dataset, which includes project demographics and operational metrics, human factors, organizational context, technical aspects, and external influences. Moreover, a comprehensive data reliability testing is conducted through pre-processing methods for categorical attributes, one-hot encoding, and Min-Max normalization of budget and timeline, and risk metrics. Advanced feature engineering uses graph-based feature relationships to identify hidden project attribute dependencies, Graph Signal Processing to create project attribute dependencies, and LASSO with polynomial feature expansion to achieve optimal results. The proposed TAM-Lite architecture integrates TabNet, a mini autoencoder, and a shallow multilayer for project risk prediction. Moreover, stage-wise training is conducted based on Gradient Boosted Rule Sets with Extreme Learning Machines and fuzzy logic classification. The model generates risk level probabilities, which are evaluated through Bayesian Networks and counterfactual explanations to deliver clear and actionable risk reduction recommendations.

DOI: https://doi.org/10.5281/zenodo.19483001

Explainable Artificial Intelligence (XAI)System For Machine Learning Decisions

Authors: Ms. Gyara Monika, Banothu Malsoor, Mendu Balram Sai Abhishek, Mohammed Abdul Sameer

Abstract: Explainable Artificial Intelligence (XAI) is a system that helps humans understand how machine learning models make decisions. Traditional AI models often work like a “black box,” where the output is given without explaining the reason. XAI provides clear explanations for predictions by showing important features, rules, or visual insights. This improves transparency, trust, and fairness in AI systems, especially in critical areas like healthcare, finance, and education. By making AI decisions understandable, XAI helps users and developers detect errors, bias, and improve model performance.

Semantic And Contextual Intelligence-Based Court Verdict Prediction

Authors: Mr. Chitoor Venkat Rao Ajay Kumar, Amgothu Shivateja, Dhanavath Praveen, Banoth Bhaskar

Abstract: The “Semantic and Contextual Intelligence-Based Court Verdict Prediction” system uses AI to analyze legal case data and predict outcomes. It understands the meaning and context of legal documents using Natural Language Processing. The system studies past judgments, case facts, and legal patterns to make predictions. It provides structured insights that help legal professionals in decision-making. This improves the speed, accuracy, and efficiency of legal analysis.

DOI:

Cyberthreats Information In Real-time

Authors: Mrs. Kalluri Jaya Sri Sai, Dheeravath Rajender, Bommapala Manideep, Arishe Pramod

Abstract: With the increasing demand for advanced digital security, efficient and scalable real-time monitoring has become essential. Traditional security evaluation methods often rely on manual oversight or delayed reporting, which lacks the immediate and personalized feedback necessary to thwart modern attacks. This project presents an Intelligent System for Real-Time Cyberthreat Information that leverages automated data streaming to evaluate the digital landscape for threats. The proposed system analyzes network logs and global threat feeds for syntax, logic, and patterns of malicious activity, providing instant alerts along with clear threat explanations and suggested mitigation strategies.

DOI:

Medical Image Analysis Tool: An Ai-Powered Diagnostic Assistant For Medical Imaging

Authors: K.Satheesh, S.Chandana, V.Balaji, Sk.Yaseen, V.V.Bhanu Satya Sri

Abstract: The Medical Image Analysis Tool uses Google’s Gemini 2.5 Flash multimodal model to analyze X-rays, MRIs, CT scans, and ultrasound images, generating comprehensive diagnostic reports with findings, diagnoses, and patient-friendly explanations. Integrated with DuckDuckGo search, it enables real- time retrieval of medical literature and treatment protocols for evidence-based recommendations. Built with Streamlit and the Agno framework, the tool delivers structured, medically accurate responses in markdown format for healthcare professionals and students. This AI-powered assistant reduces diagnostic uncertainty, empowers data- driven decision-making, and enhances medical image interpretation efficiency.

Carbon Purification System

Authors: Dr. M. S. Yadhav, Mrs. S. V. Zanjad, Abhishek Prakash Lohar, Malhar Ravindra Kale, Vivek Surendra Gadekar, Avinash Mariba Paikrao.D

Abstract: The Carbon Purification System is designed to improve the quality of gas produced during the decomposition of organic waste. Biogas generated from kitchen waste or other biodegradable materials contains useful methane gas along with unwanted impurities such as hydrogen sulfide, carbon dioxide, and bad odor. These impurities reduce the efficiency and usability of the gas. Therefore, purification of biogas is necessary before it can be used for practical applications. This project focuses on developing a simple and cost-effective carbon purification system that uses activated carbon as the main filtering material. Activated carbon has a very large surface area with many tiny pores that can absorb harmful gases and impurities through the process of adsorption. In this system, the raw gas produced from the digester passes through different filter layers such as a pre-filter, activated carbon layer, and cotton layer, which help remove dust particles, toxic gases, and unpleasant smell. The purification chamber is designed using simple materials so that it can be easily implemented in small-scale applications such as homes, laboratories, and small biogas plants. As the gas passes through the filter layers, harmful substances are trapped and the output gas becomes cleaner and safer to use.

DOI: https://doi.org/10.5281/zenodo.19492379

Mobile Phone Detection System Using ESP32, HMC5883L, NRF24L01, LCD Display, and Buzzer

Authors: Vishva Shedge, Shubham Lashkar, Swaraj Pawar, Aaditya Shinde, Prof. A. N. Dubey

Abstract: This paper presents the design and implementation of a Mobile Phone Detection System intended for deployment in restricted environments such as examination halls, secure meeting rooms, and classified zones. The proposed system integrates an ESP32 microcontroller with an NMC5883L digital compass module (HMC5883L-compatible) to detect the electromagnetic and magnetic field signatures associated with active mobile devices. Upon detection, the system triggers an audible alarm via a buzzer and displays status information on a 16×2 LCD screen. Wireless data transmission using the NRF24L01 module enables communication between multiple sensor nodes and a central monitoring unit. The system is designed to be low-cost, energy-efficient, and scalable for multi-zone surveillance. Experimental results confirm reliable detection of active mobile phones within a defined proximity range, demonstrating the practical viability of the proposed approach.

DOI:

Review Paper on Experimental Investigation on Partial Replacement of Cement by (Ggbfs) and Partial Replacement of Course Aggregate By Rubber Pallets

Authors: Saurabh.D. Kamble, Dr. V. P. Varghese, Prof. M.N. Umare

Abstract: The high consumption rate of raw materials by the construction sector, results in chronic shortage of building material and the associated environmental damage. In the last decade, many research on the utilization of waste products in concrete in order to reduce the utilization of natural available resource have been undertaken. Thus, in this paper we have tried to review the use of waste products which are replaced as cement partially by ground granulated blast furnace slag (GGBFS) and coarse aggregate by rubber pallets. The aim is to study literature review & determine how GGBFS & Rubber pallets would affect the compressive strength of concrete when used in different proportions of GGBFS 20%, 30% & 50%, & Rubber pallets 5%, 10%, 15%. This literature review investigation supports the potential use of GGBFS and rubber pallets as partial replacement in concrete production, contributing to eco-friendly and resource efficient construction practice.

Intelligent Web-Based System For Automated Code Assessment And Learning

Authors: Dr. CH. Kishore Kumar, Joruka Vigneshwar, SK Khaja Mohinuddin Pasha, Karingula Dileep Goud

Abstract: The “Intelligent Web-Based System for Automated Code Assessment and Learning”, designed to enhance programming education using Artificial Intelligence and Machine Learning techniques. The system allows users to submit programming code through a web interface, where it is automatically evaluated for syntax, correctness, logic, and efficiency. Unlike traditional manual evaluation methods, this system provides instant and meaningful feedback by analyzing errors, identifying logical mistakes, and suggesting improvements and optimized solutions. This enables learners to better understand their mistakes and improve their coding skills effectively. The web-based platform supports real-time code execution and evaluation, making it scalable and accessible to a large number of users. It also includes features such as performance analysis, scoring mechanisms, and personalized learning recommendations based on user performance. The system can be extended to support multiple programming languages and adaptive learning paths. Overall, this project focuses on developing a smart and efficient solution that bridges the gap between theoretical learning and practical coding skills, offering benefits such as reduced instructor workload, faster evaluation, improved learning outcomes, and enhanced user engagement through intelligent feedback.

Smart Loan: A Risk-Aware and Explainable Loan Eligibility Prediction System Using Machine Learning

Authors: Mr. R.Rajesh, Bura Keerthi, K.Keerthan Reddy, A.Siddartha

Abstract: Smart Loan is an intelligent system designed to predict loan eligibility and assess risk using machine learning techniques. Traditional loan approval processes are time-consuming and prone to human bias. This system automates the evaluation process by analyzing applicant data such as income, credit history, employment status, and financial behavior. The model predicts whether a loan should be approved and categorizes applicants based on risk level (low, medium, high). The system ensures faster decision-making, reduces default risks, and improves efficiency for financial institutions.

Smart Qr Code and Geo-Fenced Attendance System

Authors: Muthulakshmi M, Saravanan P, Srihari M, Shunmugapandian P

Abstract: This paper presents Q-Track, a Smart QR Code and Geo-Fenced Attendance System designed to provide a secure, efficient, and automated solution for attendance management in educational institutions. Traditional attendance systems, including manual registers and biometric methods, suffer from limitations such as time consumption, proxy attendance, and lack of real-time monitoring. To overcome these challenges, the proposed system integrates dynamic QR code generation with geo-location verification.In this system, a unique and time-bound QR code is generated for each class session by the faculty. Students scan the QR code using their mobile devices to mark attendance. To ensure authenticity, the system incorporates geo-fencing technology, which validates the real-time location of the student. Attendance is recorded only when both QR authentication and location verification are successful, thereby eliminating proxy attendance and ensuring reliability.The system is implemented as a web-based application with a user-friendly interface accessible on both mobile and desktop devices. It includes modules for user authentication, QR code generation, attendance tracking, and report generation. Real-time data processing enables faculty to monitor attendance instantly and generate detailed reports for analysis.The proposed solution enhances accuracy, reduces manual workload, and improves transparency in attendance management. By combining QR technology with geo-location services, the system provides a scalable and cost- effective approach suitable for modern academic environments.

AI Tool/mobile App For Indian Sign Language(ISL) Generator From Audio Visual Content In English/Hindi To ISL Content And Vice-versa

Authors: Dr. Harsha R. Vyawahare, Sukhada Shripad Tare, Ashwini Nitin Shingane, Shreya Sunil Shinde, Bhavika Suraj Jain

Abstract: This paper presents a practical and lightweight bidirectional communication system that translates between speech/text and Indian Sign Language (ISL) using machine learning and computer vision techniques. The system supports two modes: Speech-to-ISL and ISL-to-Text/Speech. In Speech Mode, spoken input is converted into text using speech recognition, then mapped to corresponding ISL alphabet images. In Camera Mode, hand gestures are captured using a webcam and classified using a Convolutional Neural Network (CNN) model to generate text and voice output. The system is implemented using Streamlit for the user interface, OpenCV for image processing, TensorFlow/Keras for gesture recognition, and pyttsx3 for speech synthesis. The proposed system provides a simple, real-time, and cost-effective solution to improve communication accessibility for the Deaf and Hard-of-Hearing (DHH) community

DOI: https://doi.org/10.5281/zenodo.19484295

ChatVerse: A Multilingual Chat Application For Real-Time Cross-Language Communication

Authors: Saurav Patankar, Abhijeet Waghmare, Swaraj khadhe, Durvesh Kavire, Prof. Pradnya Satpute

Abstract: Communication across different languages has become a major challenge in today’s globalized world. The need for a system that enables seamless interaction between users speaking different languages has led to the development of multilingual communication platforms. This paper presents Chat Verse, a multilingual chat application that allows users to communicate in real time without language barriers. The application provides automatic language detection and real-time message translation, enabling users to send messages in their native language while the system translates them into the receiver’s preferred language. The system is developed using Android Studio, with Java for application logic and XML for user interface design, ensuring a responsive and user-friendly experience. Chat Verse includes essential features such as user authentication, private and group chat functionality, language preference settings, notification system, and feedback module. The application focuses on delivering a smooth communication experience by integrating translation capabilities within the chat interface. The motivation behind developing Chat Verse is to create an efficient, accessible, and intelligent communication platform that removes language barriers and enhances global connectivity. Traditional messaging applications often lack seamless multilingual support, making communication difficult for users from different linguistic backgrounds. This system aims to address that limitation by providing an intuitive and automated translation-based chat environment. The proposed system emphasizes usability, efficiency, and scalability, and demonstrates how multilingual chat applications can play a significant role in improving communication in fields such as business, education, and social networking. The study also highlights the future potential of integrating advanced AI-based translation techniques for even more accurate and context-aware communication.”

DOI:

Unified Health System Using Spring Boot, MongoDB, And React JS

Authors: Nishikant Kshirsagar, Manas Lonkar, Pratik Ingle, Suraj Kushwaha, Prof. Madhavi Patil

Abstract: The Unified Health System (UHS) integrates multi- ple healthcare stakeholders into a single digital platform to im- prove patient care, records management, and treatment decision- making. This paper examines the design and implementation of a web-based UHS using Spring Boot for backend microservices, MongoDB as a NoSQL cloud database, and React JS as a dy- namic frontend framework. The system enables real-time access to patient medical history, digital prescriptions, lab reports, and appointment scheduling with hospitals. The project demonstrates reduced administrative delays, secure role-based data access, and a modern patient-centric healthcare experience. Existing research confirms that fragmented healthcare data and the absence of interoperable systems remain critical barriers to efficient clinical outcomes [1], [3]. By adopting a microservice-based approach and leveraging NoSQL document storage, this system overcomes the scalability limitations of monolithic architectures. The find- ings align with recent studies demonstrating that cloud-based digital platforms can significantly enhance healthcare workflow efficiency and reduce manual intervention [5], [7].

DOI:

Twitter Sentiment Analysis Using BERT: A Transformer-Based NLP Approach

Authors: M.S.R.naidu, Barri Kuvalaya, Bandaru Jyothika, Barle hemanth kumar, Amjuru bhanuprakash

Abstract: This paper introduces Bidirectional Encoder Representations from Transformers (BERT), a transformer-based natural language processing framework for sentiment analysis of Twitter data. Large amounts of opinion-rich textual data are produced by social media platforms, reflecting the public’s feelings about societal issues, events, and products. Conventional sentiment analysis methods have trouble deciphering the informal language, contextual meaning, and semantic ambiguity seen in tweets. A pretrained BERT model is optimized for multi-class sentiment classification in order to get over these restrictions. An end-to-end pipeline comprising data preprocessing, tokenization, model training, evaluation, and result display is demonstrated in the built notebook. Experimental data reveal that contextual embeddings and attention mechanisms greatly boost sentiment classification accuracy compared to conventional approaches, validating the usefulness of transformer-based models for social media opinion mining.

Fake News Detection

Authors: Adlin Jebakumari, Uzefa Begum, Kathi Harshitha Reddy, Mohammed Rameez

Abstract: The growth of digital media in recent years has created a major public issue. This is evident in the increase of false information, often called fake news. Fake news refers to any news item that contains false information for the audience. This research project combines traditional machine learning methods with modern deep learning techniques to detect fake news using a hybrid detection system. The news articles will undergo several preprocessing steps: text cleaning, tokenization, stop word removal, and text data normalization for analysis. The team will preprocess the textual data, which will then be converted into numeric data for machine learning and deep learning models. This will use feature extraction methods like tokenization and word embeddings. The project will apply traditional machine learning models to create training data that captures the unique features of fake news and real news articles. The study will also use various deep learning models, including LSTM Networks and BERT. These models will help identify sequential and contextual relationships in articles by understanding complex language patterns and the connections among different types of text data.

DOI: https://doi.org/10.5281/zenodo.19508418

Blockchain Based Certificate Management And Verification System

Authors: Jayashree Pasalkar, Vedant Mahanavar, Pranav Patil, Om Mahajan

Abstract: Counterfeit academic certificates have increased sig- nificantly enough so they now create problems for many organiza- tions (i.e., schools, employers, government agencies) because they reduce faith in the ability of organizations to verify credentials. Most current methods used to manage academic certificates are primarily manual and/or based on centralized database storage; therefore, most are subject to various forms of manipulation (e.g., unauthorized access/modification), delayed processing, and additional risks associated with verification processes. Blockchain technology has recently emerged as a possible solution for authenticating certificates securely; however, many of the current blockchain implementations are built upon platforms such as Ethereum, which experience both high transaction costs, and limited scalability. To overcome these constraints, this research will present a blockchain-based certificate management and verification system that utilizes the high-performance and low cost attributes of the Solana blockchain platform with a Django- based backend system. With this system, academic institutions can issue certificates (while maintaining the original formatting), or register external certifications issued to students/alumni. All generated certificates are hashed using the SHA-256 hashing algorithm, and each unique hash is stored on the Solana blockchain via a Rust-based Anchor smart contract. Upon receipt of a certificate to be verified, the proposed system hashes the submitted certificate, and then compares its hash value with the unalterable blockchain record to authenticate/verify the legitimacy of the submitted certificate, or identify if the submitted certificate was altered/tampered. In combination with the security provided by blockchain, the scalability of the Solana blockchain, and an efficient backend architecture, this proposed system provides a highly effective method of verifying the authenticity of academic certificates, while reducing the risk of fraudulent activity.

DOI: https://doi.org/10.5281/zenodo.19508445

Agentic AI-Based Interview Preparation Assistant

Authors: Shashank Tiwari, Amrutha Uppala, Manasa Aerragunta, Rishikar Ummadi

Abstract: — Interview preparation is an important process for students and job seekers, but traditional preparation methods often lack personalized feedback and real interview experience. In this paper, an Agentic AI–based Interview Preparation System is presented that simulates interview scenarios and evaluates candidate responses. The system generates role-based interview questions using a job role and skills dataset and evaluates answers using Natural Language Processing techniques. It also provides feedback and improvement suggestions to help candidates enhance their performance. By automating interview practice and evaluation, the system provides a structured and interactive way to prepare for interviews. Overall, this approach improves interview readiness, confidence, an d skill assessment in a cost-effective and accessible way.

DOI: https://doi.org/10.5281/zenodo.19508550

NeuroFocusAI

Authors: V. Mounica, Peteti Anuneha, Shaik Abdul Karimulla, Sadhu B.S.V.V.N.S.R. Prasanth, Rangineni Sai Swarup, Badviti Sai Deepak

Abstract: Student engagement monitoring in modern classroom and online learning environments presents a significant challenge, as traditional attendance-based systems measure physical presence but fail to quantify cognitive attention. This paper presents NeuroFocusAI, an AI-based student concentration monitoring system that evaluates real-time attention levels using a multi-modal analysis pipeline comprising facial landmark tracking, eye gaze estimation, blink detection, emotion recognition, and environmental noise analysis. The system processes live webcam input using the MediaPipe FaceMesh model, which detects 468 facial landmark points to enable precise iris-based gaze tracking and Eye Aspect Ratio (EAR) blink detection. Emotional state classification is performed using the DeepFace library across six emotion categories. Environmental noise levels are concurrently measured using Root Mean Square (RMS) audio signal processing via the SoundDevice library. A weighted scoring algorithm combines gaze direction (60%), emotion state (20%), and environmental noise (20%) to compute a concentration score between 0 and 100, which is stored periodically for session analytics. The backend is implemented using FastAPI, with SQLite as the persistent data store, and a React.js-based dashboard provides real-time analytics for both students and teachers. Experimental results demonstrate that the system accurately classifies student attention into three levels — High Focus (80–100), Moderate Focus (60–79), and Low Focus (0–59) — with significant improvements over traditional attendance-based engagement measurement.

DOI:

Smart Campus Energy Usage Analysis And Prediction

Authors: sasiram anupoju, Lakshmi Narasimham gorthi, sai kalyan nallamadhi, suhas rallabandi

Abstract: This project presents the design and development of a Smart Campus Energy Usage Analysis and Prediction system that monitors and forecasts energy consumption across campus facilities. The system collects energy usage data from different buildings such as hostels, academic blocks, and libraries, and processes it using data analytics and machine learning techniques. A predictive model based on linear regression is used to estimate future energy consumption patterns. The system also provides interactive dashboards for real-time visualization, including consumption trends, building-wise distribution, and forecast insights. The proposed system aims to improve energy efficiency, reduce wastage, and support sustainable energy management in smart campus environments.

DOI: https://doi.org/10.5281/zenodo.19509795

Fit Fuel : Fuel Your Body, Train Smarter

Authors: Dr. CH. Kishore Kumar, Vovaldas Tejaswini, Beeram Pranaya, Diya Shaik, Sana Shaik

Abstract: Fit Fuel is an integrated web-based fitness and nutrition platform designed to help users exercise correctly and maintain healthy eating habits in one place. Unlike fragmented solutions that separate workout guidance and diet planning across multiple platforms, Fit Fuel unifies both services within a single website for better convenience, consistency, and personalization. The system provides muscle- specific exercise guidance using clear posture images that help users understand correct workout techniques without relying on video streaming. In addition to exercise guidance, Fit Fuel generates personalized Indian meal plans based on the user’s daily calorie requirements, dietary preferences, and allergies. The platform also includes features such as streak tracking, daily journaling, and profile management to encourage regular engagement and long- term habit formation. By combining fitness instruction, nutrition planning, and motivational tools into a single web interface, Fit Fuel promotes a holistic and user-friendly approach to health management.

DOI: https://doi.org/10.5281/zenodo.19509880

Crowd Aware Public Space Monitor

Authors: Adi Gowri Tejaswini, DJ Rishika, Rumaan Tamheen, Vasa Sravya

Abstract: Monitoring crowd density is a crucial task for ensuring safety and preventing overcrowding-related issues. The traditional methods for monitoring crowds involve manual observation and camera surveillance, which are time-consuming and require continuous monitoring. This paper proposes a hybrid approach for crowd detection using Raspberry Pi, incorporating wireless device detection, Bluetooth scanning, infrared sensing, and computer vision. The system estimates the crowd density based on wireless device detection and verifies the presence of people through OpenCV-based human detection. The infrared sensor is used to improve the accuracy of the system by tracking entry and exit movements. The hybrid approach is an improvement over traditional methods, reducing the limitations associated with each method. The paper also discusses different approaches to crowd detection, highlighting the advantages and limitations of these methods, and the benefits of a hybrid approach for real-time applications.

DOI: https://doi.org/10.5281/zenodo.19510228

A Quantum-Edge Deep Reinforcement Learning Framework For Adaptive And Privacy-Preserving Dynamic Pricing In E-commerce

Authors: Mr. Akula Sri Naga Sai Veera Pawan Anirudh, Mrs. G Prameela

Abstract: The rapid rise of e-commerce platforms has created a need for complex pricing systems that react to market conditions in real-time to improve market share and customer satisfaction. In this paper, we present a new Edge-AI powered situational pricing optimization framework based on a Deep Reinforcement Learning (DRL) model, leveraging the low latency pricing decision-making capability of a distributed edge computing network. In our model, we use federated learning processes with multi-agent deep reinforcement learning to create hybrid pricing intelligence based on the ongoing analysis of patterns of customer behaviour, competitors and market volatility signals. Our framework offers a solution to the fundamental limitations of cloud-based traditional pricing systems (and understandings) in shipping complex processes to ultra-sophisticated AI pricing engines that function on lightweight AI models located at edge nodes in the network, improving latency from seconds to milliseconds. Our experimental validation based on real e-commerce data shows a 23.4% im-provement in revenue optimizations, 18.7% improvements in reduction for de-cision latency of price adjustments and a remarkable 31.2% increase in customer satisfaction metrics relative to the previous centralized mode (cloud-based). This system offers a decentralized framework that can scale globally to support multi-market e-commerce operations, while also improving data privacy and confidential processing in compliance with regulatory demands.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.178

A Hybrid Framework For Real-Time Android Malware Detection Using Machine Learning And Deep Learning

Authors: P. Chakradhar Rao, Vakadi venkata krishna

Abstract: The study proposes an efficient and secure hybrid framework for detecting Android malware in modern mobile environments. The widespread adoption of Android smartphones has led to increased security risks, as these devices are frequently targeted by sophisticated malware attacks. Furthermore, the growing integration of Android applications with Internet of Things (IoT) systems amplifies the potential impact of such threats. Detecting malware manually in large-scale and continuously evolving datasets is both time-consuming and ineffective. To address these challenges, our approach integrates real-time data acquisition and deep learning techniques. Malware hash values are dynamically updated using data extracted from Twitter at regular intervals of 48 hours, ensuring the system remains up-to-date with emerging threats. In addition, application features, particularly permissions, are analyzed using a Recurrent Neural Network (RNN) with Long Short-Term Memory (LSTM) architecture for accurate classification. The model is trained and evaluated to distinguish between benign and malicious applications, achieving a detection accuracy of approximately 94%. The proposed multi-layer framework enhances detection efficiency by combining traditional signature-based methods with intelligent learning mechanisms. This integrated system improves reliability, strengthens mobile security, and provides an effective solution for real-time Android malware detection and prevention.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.177

Hybrid Machine Learning Approach For Fishermen Safety And Communication In Marine Environments

Authors: M. Sujana Priyadarshini, Gunduprolu vijayakumar

Abstract: The study proposes an intelligent and reliable hybrid framework for enhancing fishermen safety and communication in marine environments using machine learning and electromagnetic water networks. Fishing activities in deep-sea regions involve significant risks due to unpredictable weather conditions, accidental border crossings, and limited communication facilities. Traditional monitoring systems rely heavily on manual observation and basic GPS tracking, which are often inefficient in handling real-time emergencies and dynamic ocean conditions. Additionally, the lack of continuous monitoring and predictive capabilities increases the vulnerability of fishermen to accidents and environmental hazards.To address these challenges, the proposed system integrates real-time data acquisition from multiple sources, including GPS tracking, environmental sensors, and electromagnetic sensors, to ensure continuous monitoring of marine conditions. The system employs machine learning techniques such as anomaly detection algorithms to identify abnormal vessel behavior, including sudden stops, unusual movements, and route deviations that may indicate distress situations. Furthermore, time-series data collected from sensors is analyzed using advanced deep learning techniques to predict environmental changes such as weather fluctuations and sea conditions.The model is trained and evaluated to accurately detect potential risks and provide early warning alerts, thereby enabling proactive decision-making. The proposed multi-layer framework enhances system performance by combining real-time monitoring, anomaly detection, and predictive analysis. This integrated approach improves communication between fishermen and coastal authorities through wireless technologies, ensuring timely response during emergencies.The system significantly enhances maritime safety, reduces the risk of accidents, and improves operational efficiency. By leveraging machine learning and real-time data processing, the proposed solution provides a scalable, efficient, and intelligent framework for ensuring the safety and security of fishermen in modern maritime environments.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.176

A Hybrid Privacy-Preserving Spam Detection Framework Using Machine Learning And Cryptographic Techniques

Authors: M. Sujana Priyadarshini, Akula Swathi

Abstract: The exponential growth of email communication has led to an increase in unsolicited and potentially harmful spam messages, posing significant challenges to both users and organizations. Traditional spam detection techniques primarily focus on classification accuracy while often neglecting data security and privacy concerns. This paper presents a secure and efficient email spam detection system that integrates machine learning with cryptographic techniques. The proposed approach utilizes Support Vector Machine (SVM) for effective classification of emails based on textual features. To ensure data confidentiality, Advanced Encryption Standard (AES) is employed for encrypting email content, while Elliptic Curve Cryptography (ECC) is used for secure key exchange. The integration of classification and encryption mechanisms enables the system to provide reliable spam detection while preserving sensitive information. The proposed framework is suitable for real-world applications where both accuracy and data privacy are essential.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.175

Conversational Product Recommendation System Using LLM

Authors: Mr. Shashank Tiwari, Kommunuri Ashok Kumar, Valigonda Laxmiprasanna, Kayitha Sai Rachana

Abstract: The Conversational Product Recommendation System using LLM is an AI-driven application designed to enhance product recommendation by enabling natural language interaction between users and the system. In modern e-commerce environments, users often face information overload due to the vast number of available products. Traditional recommendation systems rely on static filtering methods and fail to understand complex user queries expressed in natural language. To address these limitations, the proposed system integrates Large Language Models (LLMs) with Natural Language Processing (NLP) techniques to interpret user intent, preferences, and constraints. The system provides a chatbot-based interface where users can interact conversationally, refine their queries, and receive personalized product recommendations in real time. A recommendation engine processes extracted features and ranks products based on relevance, while a backend database manages product data and user interactions.

Optimization Of Transformer Design Parameters Using Altair Flux

Authors: K V Bharadwaj Karthik, V. Abhilash Naik, V Kowshik Chavali, G Suresh Babu

Abstract: This project focuses on the analysis of a three-phase star–delta step-down transformer using Altair Flux with emphasis on the no-load test and short- circuit test. The objective is to accurately evaluate core (iron) losses and copper (Joule) losses through finite element electromagnetic simulation. In the no-load test, rated voltage is applied to the primary winding while the secondary is kept open, enabling determination of magnetizing current, flux distribution, and core losses. The Bertotti loss model is employed within Altair Flux to separate hysteresis, eddy current, and excess losses in the core. Flux density distribution is examined to ensure operation below saturation limits. In the short-circuit test, the secondary winding is shorted and a reduced voltage is applied to circulate rated current, allowing evaluation of winding resistance, leakage reactance, and copper losses. The simulation accurately captures current density and I²R losses in both primary and secondary windings. The star–delta connection is properly modeled to obtain correct phase relationships and loss values. Results from Altair Flux demonstrate realistic loss estimation consistent with transformer theory. The study confirms that core losses remain nearly constant with load, while copper losses vary with the square of current. Overall, the work validates Altair Flux as an effective tool for detailed electromagnetic analysis of transformer performance using standard no-load and short-circuit test procedures.

An Efficient XGBoost-Based Approach For Electric Load Forecasting In Smart Energy Systems

Authors: Dr. P.Vamsi krishna raja, Nama Venkata Bhaskara Sudheer

Abstract: Electric load forecasting plays a crucial role in efficient power system operation and energy management. Accurate prediction of electricity demand helps in reducing operational costs and improving system reliability. However, traditional forecasting methods often fail to handle complex and non-linear patterns present in real-world data. To address this issue, this paper proposes a machine learning–based approach using Extreme Gradient Boosting (XGBoost) for electric load forecasting. The proposed system utilizes historical load data along with important features such as time and temperature to train the model. Data preprocessing and feature selection techniques are applied to improve data quality and model performance. XGBoost, a powerful ensemble learning algorithm, is employed to capture complex relationships and enhance prediction accuracy. The model is evaluated using standard performance metrics, and the results demonstrate improved accuracy and efficiency compared to conventional methods. The proposed approach provides a reliable and scalable solution for electric load forecasting, supporting better decision-making in power system planning and management.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.179

Autonomous Signal Deception and Offensive System for Battlefield Application

Authors: P. Dr. K. Rama Linga Reddy, Penumarty Srilakshmi Bhanupriya, Nikitha Mora, B. Suchitra, Sruthi Gujjula

Abstract: The paper describes the design of an Autonomous Signal Deception and Offensive System to be used at the Battlefield – a Simulink-based RF electronic warfare (EW) simulation system based on an earlier created hardware prototype, the Ultrasonic Deception System. In the previous system, four deception methods, including range deception, angle deception, stealth, and noise injection are shown with Arduino Mega and ultrasonic sensors, and the choice of the technique is done manually by an operator. The proposed system completely removes any manual involvement and adds five important extensions: autonomous selection of deception techniques with a randomized decision engine, an offensive electromagnetic pulse (EMP) generation subsystem, a cryptographic Identification Friend or Foe (IFF) protocol based on challenge-response authentication via XOR operations and pre-shared secret keys, an accurate RF channel model including path loss, propagation delay, and additive white Gaussian noise (AWGN). This system is implemerandi-based technique selection, and a Countermeasure Generation block that generates high-amplitude EMP pulses. The results of simulations show that autonomous threat classification is successful, the deployment of unpredictable deception techniques, and the possibility to quantify the degradation of the enemy system. Performance is analyzed based on six metrics such as IFF classification accuracy, deception effectiveness, SNR degradation, Shannon channel capacity, deception unpredictability entropy and system health degradation rate.

DOI: http://doi.org/

Fertilizer Spraying Machine

Authors: Rajat Vijaybahadur Singh, Prasad Balasaheb Varpe, Pranav Navnath Lokhande, Krushna Suresh Awate

Abstract: Agriculture plays a vital role in the economy, and efficient farming techniques are essential for increasing crop productivity. Fertilizer application is one of the most important processes in agriculture, but traditional methods of applying fertilizers are time-consuming, labor-intensive, and often result in uneven distribution. To overcome these problems, a solar-powered fertilizer spraying machine is developed in this project. The main objective of this project is to design and fabricate a cost-effective, eco-friendly, and efficient fertilizer spraying system that reduces manual effort and ensures uniform spraying. The machine consists of a solar panel, battery, solar charge controller, water motor pump, storage tank, nozzle, flow pipes, and a four-wheel frame. The solar panel converts sunlight into electrical energy, which is stored in the battery and used to operate the motor pump. The pump creates pressure to spray the fertilizer solution through the nozzle in the form of fine droplets. The system provides several advantages such as reduced labor, time saving, uniform distribution of fertilizers, and low operating cost due to the use of solar energy. The four-wheel structure makes the machine portable and easy to operate in agricultural fields. It is especially useful for small and medium-scale farmers. This project demonstrates the effective use of renewable energy in agriculture and contributes to sustainable farming practices. The developed machine is simple in design, economical, and capable of improving overall agricultural efficiency and productivity.

DOI: https://doi.org/10.5281/zenodo.19548929

Insider Threat Detection Using Anamoly Threat Detection

Authors: Mrs. G. Monika, B. Bindu, U. Edukondalu, K.Varshith

Abstract: Insider threat is one of the biggest problems facing organizational security since insiders are individuals with authorized access to an organization’s information assets. Organizational security solutions can only detect outsider attacks and do not perform effectively when faced with malicious behaviors or accidental acts carried out by insiders. In this research paper, a method of detecting insider threat using behavioral anomaly is outlined. This solution aims at continuous observation of user behavior such as logging on, file access and general interaction with the system resources. Machine learning algorithms are employed in modeling user behavior and alerting any deviation that can imply an act of malice.

DOI: https://doi.org/10.5281/zenodo.19549305

A Hyrid CNN-MLP Model For Diaetic Retinopathy Analysis Using Retinal Images

Authors: Mr.MD. Abdul kala, V.Krupa, M.pavani M. Hemanth sai

Abstract: Diabetic Retinopathy (DR) is a serious eye disease caused by long-term diabetes. It is one of the main causes of blindness around the globe. Early detection and prompt treatment are crucial to prevent permanent vision loss. Unfortunately, traditional diagnostic methods depend on the manual inspection of retinal fundus images by ophthalmologists. This process is time-consuming, subjective, and requires specialized skills. This project presents a Hybrid CNN-MLP Model for automated detection and classification of diabetic retinopathy using retinal images. The system combines Convolutional Neural Networks (CNN) for feature extraction and Multilayer Perceptron (MLP) for classification. The CNN component effectively captures spatial features like microaneurysms, hemorrhages, and exudates. Meanwhile, the MLP classifies these features into different levels of DR severity. The system is created using Python, TensorFlow/Keras, and Flask for online interaction. Users can upload retinal images, enter patient information, and receive real-time predictions with confidence scores, medical suggestions, and downloadable PDF reports. The system also keeps a record of patient history and provides visual analytics through graphs. This proposed model shows better accuracy, efficiency, and usability. It serves as a valuable tool for early screening and supports healthcare professionals in making decisions.

DOI: https://doi.org/10.5281/zenodo.19549841

Careen Lens

Authors: Gudimella Akhilesh, Peruri Karthik Sai, Kunburu Manikanta Reddy, Dr. Atul Kumar Ramotra

Abstract: Career decision-making among engineering students is often influenced by trends rather than a proper evaluation of individual skill sets, leading to skill–career mismatch. This project presents CareerLens, an explainable skill-based career recommendation system designed to guide students in selecting suitable academic streams and job roles. The system analyzes user-provided technical skills along with proficiency levels, maps them to predefined career requirements, and computes readiness scores to generate personalized recommendations. Additionally, it identifies skill gaps and suggests improvements to enhance career readiness. By emphasizing transparency, interpretability, and skill-driven guidance, CareerLens aims to bridge the gap between student capabilities and evolving industry demands.

DOI: https://doi.org/10.5281/zenodo.19550142

AI Framework For Personalized Fitness & Diet Recommendation System

Authors: Ranjith Durgunala, Harshith Manchikkanti, Rahul Perugu, Sunadh Rithvik Ponnuru

Abstract: Therapid increase in sedentary lifestyles and unhealthy dietary habits has raised serious concerns regarding physical fitness and overall well-being. This project presents an AI Framework for Personalized Fitness & Diet Recommendation System designed to provide intelligent and customized health guidance. The system gathers essential user information including age, gender, height, weight, activity level, medical conditions, dietary preference, and fitness goals. Using this data, Body Mass Index (BMI) is calculated to assess the user’s health status. Machine learning algorithms analyze user profiles to generate personalized workout routines and diet plans tailored for fat loss, muscle gain, weight gain, or general fitness. A progress tracking module records daily weight, workout completion, and calorie intake to evaluate improvement. In addition, predictive models estimate expected fitness outcomes over 30, 60, and 90 days. The proposed framework enhances decision-making through data-driven insights, improves user engagement, and promotes sustainable lifestyle changes using artificial intelligence and machine learning techniques.

Building Responsible AI Tools For Small Scale Business.

Authors: Md Ali Bashar Alam, Uzma Fathima, Dr.A. Kannagi

Abstract: The rapid adoption of artificial intelligence (AI) has created significant opportunities for innovation, efficiency, and competitive growth among small-scale businesses. However, limited resources, lack of technical expertise, and growing ethical concerns make it challenging for small enterprises to implement AI responsibly. This research paper explores the design and development of responsible AI tools tailored specifically for small-scale business environments, focusing on transparency, fairness, accountability, data privacy, and regulatory compliance. By analysing existing global AI ethics frameworks and governance principles, the study proposes a practical model that integrates ethical guidelines into scalable and cost-effective AI solutions. The research highlights key challenges such as algorithmic bias, data protection risks, limited infrastructure, and information asymmetry faced by small businesses, while presenting strategies to mitigate these issues through explainable AI, lightweight governance mechanisms, and certification-based approaches. Furthermore, the paper discusses how responsible AI adoption can enhance customer trust, reduce reputational risk, and support sustainable digital transformation. The findings aim to bridge the gap between high-level ethical principles and real-world implementation by offering a structured framework that enables small-scale enterprises to deploy AI systems safely, ethically, and efficiently. Ultimately, this study contributes to the advancement of inclusive and trustworthy AI ecosystems by empowering small businesses to adopt responsible innovation practices without compromising operational feasibility or economic growth.

AI-based PCOS Anemia Early Risk Detector

Authors: Gayathri Kodipaka, Kompalli Sri Divya Muktha, Sowmya Manukonda

Abstract: Polycystic Ovary Syndrome (PCOS) and Anemia are among the most prevalent yet underdiagnosed health conditions affecting women in India, largely due to delayed symptom recognition, lack of awareness, and limited access to preventive healthcare. This project presents an AI-based early risk detection system designed to provide non-diagnostic risk assessment and health awareness support. The system analyzes user-provided inputs such as lifestyle habits, menstrual irregularities, fatigue levels, dietary patterns, and basic lab values like hemoglobin range to estimate a personalized risk probability for PCOS and Anemia. Machine learning models including Logistic Regression and XGBoost are employed to identify patterns associated with elevated risk levels. The application is developed using Python for model implementation, Streamlit for an interactive and accessible user interface, and SQLite for lightweight data storage. Unlike conventional period-tracking applications, this solution focuses on preventive risk scoring tailored to Indian women, aiming to encourage early medical consultation and improve health outcomes across both rural and urban populations.

DOI: https://doi.org/10.5281/zenodo.19552421

Comprehensive Technical Analysis Of Nuclear Thermal And Nuclear Electric Propulsion Systems For Interplanetary Exploration

Authors: Aashutosh Kushwaha, Tapas Kumar Nandi

Abstract: The advancement ofi human civilization into the solar system is fiundamentally constrained by the energy density limitations ofi chemical propulsion. Nuclear propulsion, encompassing thermal, electric, and pulse architectures, ofifiers a transfiormative leap in specifiic impulse and payload capacity by leveraging the high energy density ofi nuclear fiission. This report provides a technically rigorous examination ofi the evolution, physics, and design ofi nuclear rocket systems. It begins with a detailed historical reconstruction ofi the United States’ Project Rover and NERVA programs, alongsidg thg Sovigt Union’s RD-0410 development, highlighting the achievement ofi specifiic impulses exceeding 840 seconds. The fiundamental physics ofi neutron kinetics and heat transfier in extreme environments are derived, fiocusing on the McCarthy- Wolfi and Taylor correlations fior supercritical hydrogen. A comparative analysis ofi propellants—liquid hydrogen, ammonia, and methane—reveals the critical trade-ofifis between mass efifiiciency and storage density. Advanced concepts, including gas-core reactors, nuclear light bulbs, and the pulse propulsion ofi Project Orion, are evaluated fior their potential to achieve interstellar velocities. The report concludes with an analysis ofi the current DARPA/NASA DRACO mission and the shifit toward High-Assay Low-Enriched Uranium (HALEU) fiuels, outlining a path fior the next generation ofi deep-space transportation.

DOI: https://doi.org/10.5281/zenodo.19553037

Web based Smart City Compliant Monitoring and Resolution System

Authors: Assistant Professor K.Karthick, R.Mounish, M.Sanjay, M.Vallarasu

Abstract: A Web-Based Smart City Complaint Monitoring and Resolution System is an integrated digital solution developed to enhance the efficiency of urban service management. The system provides a user-friendly web interface that enables citizens to register complaints related to public services such as infrastructure, sanitation, water supply, electricity, and traffic management. Each complaint is automatically classified, assigned a unique identification number, and routed to the appropriate municipal authority for prompt action. The proposed system incorporates real-time tracking, allowing users to monitor complaint status throughout its lifecycle. Administrative modules support prioritization based on severity, geographic location, and impact, thereby improving response time and service delivery. A centralized database facilitates data storage, analysis, and reporting, enabling authorities to identify recurring issues and support data-driven decision-making for smart city development. Furthermore, features such as automated notifications, feedback mechanisms, and performance dashboards ensure transparency, accountability, and improved citizen engagement. The implementation of web-based technologies minimizes manual intervention and enhances communication between stakeholders. Overall, the system promotes efficient governance and contributes to the development of sustainable and citizen-centric smart cities.

DOI: http://doi.org/

Mythological Motifs In The Master And Margarita An Intertextual And Symbolic Analysis

Authors: Ekaterina Nikiforova

Abstract: This paper conducts an in-depth exploration of the mythological motifs in The Master and Mar-garita, focusing on the novel’s multilayered character system. It examines how Mikhail Bulgakov integrates mythology, religion, and literary traditions to construct a narrative rich in intertextuali-ty and symbolic meaning. The analysis centers on the characters of the Master, Margarita, Woland, and his retinue, revealing their deep intertextual connections with classic texts such as the Bible and Faust. It argues that these mythological motifs play a crucial role in shaping the characters, advancing the plot, and expressing philosophical themes. Through close textual read-ing and cultural comparison, this study shows that The Master and Margarita is not merely a work of fantasy but a literary masterpiece that reflects Bulgakov’s profound contemplation on good and evil, faith, freedom, and Soviet reality. This research aims to deepen the understanding of the relationship between mythology and character construction in Bulgakov’s work and to provide new interpretive perspectives on its literary and cultural significance.

DOI: https://doi.org/10.5281/zenodo.19554233

Wireless Data Transfer Using Li-Fi Technology{Text&Audio}

Authors: H. M. Pawar, Ajay Athare, Sujal Bagdane, Damini Barde, Rutika Vadaje

Abstract: In Recent Years, The Rapid Growth Of Wireless Communication Technologies Has Led To Congestion In The Radio Frequency (Rf) Spectrum. Technologies Such As Wi-Fi, Bluetooth, And Cellular Communication Rely Heavily On Rf Waves, Which Suffer From Limitations Like Limited Bandwidth, Interference, Security Issues, And Restricted Usage In Sensitive Environments Such As Hospitals And Aircraft. To Overcome These Challenges, Light Fidelity (Li-Fi) Has Emerged As A Promising Alternative Wireless Communication Technology.

DOI: https://doi.org/10.5281/zenodo.19554367

Enhancing Workforce Performance Through Digital HR Transformation: An Empirical Study Of Employee Productivity And Work Engagement

Authors: Dr. B. Mohan Kumar, Dr. Y.S.S. Patro, Ms. J. Lavanya

Abstract: Digital Human Resource Transformation (HRDT) is a key driver of workforce performance in the digital era. This study analyzes the impact of digital HR practices on employee productivity and work engagement using data from 140 respondents. Statistical techniques such as correlation and regression were applied. The results show a significant positive relationship between HRDT, work engagement, and productivity, with engagement partially mediating this relationship. The study highlights the importance of technology-enabled HR systems in enhancing employee outcomes and organizational effectiveness.

DOI: https://doi.org/10.5281/zenodo.19554455

A Review On Insulin Pump Therapy

Authors: Khushi V. Kayande, Kishor B. Charhate, Kiran. B. Nagre, Krushna S. Bhutekar, Dr. Prafulla R Tathe, Kishor B. Charhate

Abstract: Insulin pump therapy, clinically known as Continuous Subcutaneous Insulin Infusion (CSII), represents a sophisticated shift in diabetes management from conventional multiple daily injections (MDI). By delivering a continuous supply of rapid-acting insulin through a subcutaneous cannula, the therapy more closely mimics the physiological insulin secretion of a healthy pancreas. The current data indicate there are over 1 million people with diabetes on insulin pump therapy worldwide3 and 350,000 to 515,000 in the United States4. Insulin pump therapy offers increased lifestyle flexibility and improved glucose management. The goal of this paper is to outline the topics that should be covered by diabetes care and education specialists when teaching people with diabetes (PWD) and their families or significant others. It focuses on insulin pump therapy and the importance of maintaining a high level of expertise in this subspecialty of diabetes education if choosing to include pump and sensor training in the individual specialist’s practice.

DOI: https://doi.org/10.5281/zenodo.19554544

Lumière-Café-Management

Authors: Atharva Deshmukh, Anirudh Madiwal, Shruti Ayare, Purva Devrukhkar

Abstract: The rapid growth of the food and beverage industry has increased the need for efficient and automated management systems. Traditional café management methods often rely on manual processes, leading to inefficiencies, errors, and reduced customer satisfaction. This research paper presents the design and development of the Lumière Café Management System, a full-stack web-based application aimed at streamlining café operations including order processing, inventory tracking, staff management, and reporting. The system leverages modern web technologies such as React for the frontend, Node.js with Express for backend services, and SQL-based databases for persistent data storage. The application provides role-based access for administrators and staff, enabling efficient task allocation and monitoring. The results demonstrate improved operational efficiency, reduced human errors, and enhanced decision-making capabilities through real-time data insights.

DOI:

Instagram Fake Account Detection Using Machine Learning

Authors: Himani Atul Khamkar, Riddhika Dattaram Zolage, Prof. Sanjay Eknath Gawli

Abstract: Social media platforms such as Instagram are widely used for communication, networking, and content sharing. How- ever, the rapid growth of these platforms has also led to a signifi- cant increase in fraudulent or fake accounts. These accounts are often involved in activities such as spamming, phishing, spreading misinformation, and manipulating engagement metrics. Due to the large number of users and the dynamic behavior of social media platforms, manual identification of fake accounts becomes difficult and inefficient. This research proposes a machine learning-based approach to detect fake Instagram accounts using profile-based features. Various attributes such as follower-following ratio, number of posts, engagement behavior, profile completeness, and other pro- file characteristics are analyzed. The Random Forest classification algorithm is used to distinguish between real and fake accounts. The performance of the model is evaluated using metrics such as accuracy, precision, recall, and F1-score. The experimental results demonstrate that the proposed approach can effectively identify fake accounts and contribute to improving the reliability and security of social media platforms.

DOI:

A Low-Cost Assistive Platform For Braille Learning And Accessibility In Resource-Constrained Environments

Authors: Ch. Saiharsha, D. Sathvika, J. Siddhi Haarika, P. Yuktha Laasya

Abstract: Braille literacy is very vital in facilitation of independent reading, education and communication among the visually impaired. Availability of affordable Braille learning and display systems is however low especially in environments with limited resources whereby commercial refreshable Braille displays are prohibitively expensive. In this paper, a low-cost assistive platform that is intended to aid in the learning of the Braille, visualization, and prototype of a system is presented. The offered system makes Grade-1 patterns of Braille represented in real-time, which lets the user comprehend the character mapping and sequencing. The design is focused on affordability, modularity and implementation ease, which fits the educational institutions and assistive technology development. The system has been proven to be reliable in converting textual input into relative Braille patterns. Even though the existing implementation offers a non-tactile representation, it is an effective training/validation tool of students, educators, and researchers. The suggested solution demonstrates the possibility of creating scalable and cost-efficient assistive solutions.

ElevateX: An AI-Powered Career Guidance Platform

Authors: Saif Chaudhary, Faizan Bari, Arbab Ansari, Fahim Shaikh, Prof. Anuja Kamat

Abstract: The rapid evolution of technology and the expanding digital economy have created significant complexity in career decision-making for students and fresh graduates. Traditional career counselling approaches are static, generalized, and fail to account for individual skill profiles, evolving industry demands, and personalized learning trajectories. This paper presents ElevateX, an AI-powered career guidance platform that leverages large language models (LLMs) and natural language processing (NLP) to deliver personalized career recommendations, skill gap analysis, dynamic learning roadmaps, project suggestions, resume insights, and mock interview simulations. ElevateX engages users through an intelligent questionnaire that assesses their skills, interests, and aspirations, and subsequently generates actionable guidance tailored to their profile. Experimental evaluations demonstrate high user satisfaction, improved career clarity, and measurable gains in skill awareness among participants. The platform addresses a critical gap in accessible, personalized, and data-driven career counselling for the student community.

DOI: https://doi.org/10.5281/zenodo.19566697

Plant Disease Detection Using Leaf Image

Authors: Nupur Pradip Panchal, Sakshi Sanjay Shinde

Abstract: This research report presents the development and evaluation of a deep learning-based system for the automated detection of plant diseases through leaf image analysis. Aimed at addressing the significant economic impact of crop diseases in agriculture, the project leverages a Convolutional Neural Network (CNN) model trained on the publicly available Plant Village dataset. The implemented system processes leaf images through stages of pre-processing, augmentation, and feature extraction to classify diseases in crops such as tomato, potato, and maize. The model achieved a high classification accuracy of 96.5%, with supporting precision, recall, and F1-scores all above 95%. The study successfully demonstrates the technical feasibility of using image processing and deep learning for accurate, rapid disease identification. A key innovation proposed is the deployment of this model on a mobile application, which would provide farmers with an accessible tool for early disease detection and improved crop management, thereby enhancing agricultural productivity. The report also discusses the current limitations and potential future integrations with IoT and advanced imaging technologies for broader field application.

DOI: https://doi.org/10.5281/zenodo.19567125

SkinSight: AI-Powered Skin Disease Detection

Authors: Pratiksha Shinde, Zishan Nadaf, Onkar Bhuse, Chaitanya Deshmukh

Abstract: Skin diseases constitute a major global healthcare challenge, particularly in regions with limited access to dermatological expertise. Early and accurate diagnosis is essential to reduce disease progression and associated healthcare costs. This research presents SkinSight, an advanced artificial intelligence–based clinical decision support system for automated skin disease detection using digital skin images. The proposed system is designed by aligning a real-world deployable application with state-of-the-art research methodologies reported in recent dermatology-focused deep learning literature. SkinSight integrates an advanced preprocessing pipeline for artifact removal and illumination normalization, a two-stage validation framework to ensure input reliability, and a dual-stream deep learning ensemble combining ResNet50 and InceptionV3 architectures. Additionally, the system incorporates explainable artificial intelligence (XAI) using Grad-CAM to provide visual interpretability of predictions. A comparative analysis between the initial deployment model and the research-grade pipeline is presented, followed by a structured roadmap for bridging implementation gaps. Experimental results demonstrate that aligning deployment architecture with research-level techniques significantly improves robustness, reliability, and clinical trustworthiness. The proposed framework highlights the importance of end- to-end consistency between research and deployment in AI-driven healthcare systems.

DOI:

Steganography Hider System_928

Authors: Nitesh Baranawal, Herambh Sakpal, Pranay Manoj, Kaustabh Kadam, Prof. Mohan Kumar

Abstract: The Steganography Hider System is a secure information-hiding solution designed to protect sensitive data by embedding it within digital images, making it imperceptible to unauthorized users. Unlike traditional encryption, which only disguises data, steganography conceals the very existence of the information, providing an additional layer of security. This system employs techniques such as Least Significant Bit (LSB) substitution, transform domain methods (e.g., DCT), or advanced neural network approaches to embed secret messages while maintaining the visual quality of the cover image. The proposed system allows users to securely hide and retrieve confidential information, ensuring data confidentiality, integrity, and robustness against common image processing operations such as compression, noise addition, and format conversion. This project serves as a practical demonstration of the importance of information security in today’s digital communication era, providing a user-friendly interface that can be applied in various fields such as secure communications, copyright protection, and digital forensics.

PulseSaver- Medical Emergency Donation Network

Authors: Prof. Bhagwati Galande, Parth Bhosale, Sakshi Bathe, Aditi Gadhave, Prathamesh Amrutkar

Abstract: PulseSaver is designed to fix this. It’s an all-in-one digital network that acts as an emergency lifeline, instantly bringing together people who need help with people who can give it. We are building a single platform to manage donations of blood, organs, and money—all in real-time. We use smart AI technology to cut out the waiting. The system automatically and instantly matches the patient with the closest, most compatible, and fully verified donor or resource. Everything is secure and transparent: we rigorously verify all users and requests to build trust and ensure ethical donation practices. The main goal of PulseSaver is simple: to save lives by speeding up the emergency response and making sure critical resources get to the people who need them most, especially those in remote or underserved communities.

Smart Power Monitoring System For Home Appliances

Authors: Satyam Sharma, Jakkala Sanjay Kumar, S.Harshvardhan, Ms. Pabbu Parameshwari

Abstract: Rapid growth in residential electricity consumption has created a strong need for intelligent and user-friendly monitoring solutions that can help users understand and control their energy usage. Traditional energy meters provide only cumulative readings and do not offer real-time insights into appliance-level consumption. This paper presents the design and implementation of an IoT-based Smart Power Monitoring System for home appliances that enables continuous monitoring and remote access to electrical parameters. The proposed system measures real-time voltage, current, power, and energy consumption using ZMPT101B voltage and ACS712 current sensors interfaced with an ESP32 microcontroller. The ESP32 processes sensor data and transmits it securely to cloud platforms such as ThingSpeak and Blynk via Wi-Fi, allowing users to visualize and analyze energy usage through mobile or web dashboards. In addition, the system incorporates a relay-based protection mechanism that automatically disconnects the load during abnormal conditions such as overcurrent, voltage fluctuations, or excessive power consumption. This feature helps prevent electrical damage and improves system safety. The proposed solution is low-cost, scal-able, and energy-efficient, making it suitable for residential envi-ronments as well as small industrial applications. By providing real-time monitoring, data analytics, and remote accessibility, the system promotes energy awareness, reduces electricity wastage, and supports the development of smarter and more sustainable energy management systems.

DOI: https://doi.org/10.5281/zenodo.19569819

The Financialization Of Private Markets: Systemic Risk Beyond Banks

Authors: Vivek Sharma

Abstract: The financialization of private markets represents a fundamental transformation in the architecture of global financial systems, characterized by the growing dominance of non- bank financial institutions and the expansion of market-based credit intermediation. Over the past two decades, private markets—including private equity, private credit, hedge funds, and venture capital—have evolved from niche investment segments into systemically important components of global finance. This transformation has been significantly accelerated by post-2008 regulatory reforms, particularly Basel III, which imposed stricter capital and liquidity requirements on traditional banking institutions. While these reforms enhanced banking sector resilience, they also constrained credit supply, thereby creating space for private markets to expand and assume a central role in financial intermediation. This study investigates the systemic risk implications of the financialization of private markets, with a particular focus on the role of private credit and leveraged investment strategies. The research adopts a mixed-method approach that integrates theoretical insights from financial instability theory with empirical analysis based on secondary data. Key data sources include the International Monetary Fund (IMF), Financial Stability Board (FSB), World Bank, and industry reports from Preqin and BlackRock. The empirical framework is built around a regression model that examines the relationship between systemic risk and its primary determinants, including leverage, default rates, and market volatility. The results reveal that leverage is the most significant driver of systemic risk within private markets. High leverage amplifies both returns and losses, increasing vulnerability to adverse economic shocks and contributing to financial instability. Default rates are also found to have a strong positive relationship with systemic risk, reflecting the deterioration of borrower credit quality during periods of economic stress. Market volatility further exacerbates risk by increasing uncertainty and triggering liquidity constraints, particularly in markets characterized by illiquid assets. The study also highlights the structural vulnerabilities associated with private markets, including opacity, liquidity mismatches, and interconnectedness with the broader financial system. Unlike traditional banks, private market institutions operate with limited regulatory oversight and disclosure requirements, making it difficult to assess risk exposure and monitor systemic threats. The increasing interconnectedness between private funds, banks, and institutional investors further amplifies the potential for contagion. Overall, the findings suggest that while private markets enhance financial efficiency and provide alternative sources of capital, they also pose significant systemic risks that extend beyond the traditional banking sector. The paper concludes by emphasizing the need for expanded macroprudential regulation, improved transparency, and enhanced data reporting frameworks to ensure financial stability in an increasingly financialized global economy.

DOI: https://doi.org/10.5281/zenodo.19570015

Enhancing Employability Of ITI Qualified Students Through Strategic Management Practices

Authors: Mahe Bader Fatmi

Abstract: Industrial Training Institutes (ITIs) play a crucial role in developing a skilled workforce in India, yet many graduates face challenges in securing sustainable employment due to gaps between training and industry requirements. This paper explores how strategic management practices can enhance the employability of ITI-qualified students by aligning institutional objectives with dynamic labor market needs. It examines key strategies such as industry–institute partnerships, curriculum modernization, competency-based training, soft skills development, and the integration of digital technologies in vocational education. Drawing on selected Indian case studies, the study highlights successful models where collaboration with industry stakeholders, apprenticeship programs, and outcome-oriented training frameworks have significantly improved job placement rates. Institutions that adopted proactive management approaches—such as continuous skill mapping, faculty upskilling, and data-driven decision-making—demonstrated stronger employment outcomes for their students. The paper argues that adopting a strategic management perspective enables ITIs to transition from traditional training centers to agile, demand-driven skill hubs. It concludes by recommending policy-level support, strengthened public-private partnerships, and the institutionalization of monitoring and evaluation mechanisms to ensure long-term impact. These findings provide actionable insights for educators, policymakers, and administrators aiming to bridge the employability gap in India’s vocational education sector.

DOI: https://doi.org/10.5281/zenodo.19570413

Scrolling Into Sleeplessness: The Impact Of Social Media On Youth Sleep Pattern

Authors: Prem kumar, Aryaman Arora, Dr. A Kannagi

Abstract: The constant hum of social media has become the defining factor in the way young people and young adults sleep, or, more appropriately, the way they don’t sleep. Throughout the literature, the common factor is the way in which excessive usage of social networks, online communities, and the sheer volume of electronic communications impacts the sleep cycle in this age group. The need to remain connected at all hours has become the defining factor in sleep issues for both young people and young adults. Both biology and psychology are commonly cited as the two major factors behind sleep issues in this age group. For example, the biology of the blue light emanating from their smartphones is commonly cited as a factor in sleep issues. But more than anything, the need to remain connected has become the defining factor in sleep issues in this age group. The main aim of this paper is to examine the role of social media in the way young people and young adults sleep. The secondary aim is to identify the major factor behind the sleep issues in this age group. This will be done by using a secondary research approach to identify the major factor behind the sleep issues in this age group.

Advanced Chemistry Cell Alternatives For Electric Vehicle Battery Self-Reliance In India: A Technical And Policy Analysis

Authors: Shardul Prakash Jangam

Abstract: Achieving self-reliance (Atmanirbhar) in India’s electric vehicle (EV) sector mandates a strategic shift from reliance on imported lithium-ion (Li-ion) technology toward alternative chemistries utilizing domestically abundant raw materials. The Sodium-ion Battery (SIB) is identified as the leading alternative technology capable of ensuring immediate raw material independence by leveraging domestic reserves of sodium, iron, and manganese. SIBs are ideally suited for the cost-sensitive, short-range segments of India’s market, specifically electric two-wheeler (e-2W) and three-wheeler (e-3W) segments.1 While Li-ion currently holds a cost advantage, scaled SIB production is projected to achieve cost leadership, potentially reaching – per kilowatt-hour (kWh) at the pack level by the end of the decade.4 SIBs also offer superior safety and temperature tolerance, critical for mass EV adoption and Battery Energy Storage Systems (BESS).2 The primary recommendation is the rapid utilization of the 5 GWh capacity earmarked under the Production-Linked Incentive (PLI) Niche ACC scheme to commercialize SIBs, focusing especially on establishing a completely indigenous supply chain for the hard carbon anode using domestic biomass.7 Successful localization is projected to capture nearly 80% of the estimated 0 billion domestic EV battery market by 2030, generating substantial employment and revenue.

DOI: https://doi.org/10.5281/zenodo.19589549

A. Paperpilot: Exam Paper Generator

Authors: Aditya Phad, Shubajit Avachite, Sujal Jadhav, Rugved Gunvante

Abstract: An online auction system developed using Python provides a digital platform where users can buy and sell items through competitive bidding. The project utilizes Python for backend logic, ensuring efficient handling of user data, bids, and auction processes. It typically includes features like user registration, login authentication, and item listing. Sellers can upload product details, while buyers can place bids in real time. The system manages bid validation to ensure only higher bids are accepted. A timer mechanism is implemented to automatically close auctions after a specified duration. The highest bidder at the end of the auction is declared the winner. Data is often stored using databases such as SQLite or MySQL for persistence. The project emphasizes security, fairness, and user-friendly interaction. Overall, it demonstrates practical a pplication of Python in building scalable web-based systems.

DOI: https://doi.org/10.5281/zenodo.19589768

Iot Based Solar Wireless Power Transfer on Road for EV

Authors: Assistant Professor H. M. Pawar, Pravin Chavan, Kumar Borole, Rahul Vispute, Sona Bhalerao

Abstract: The rapid adoption of electric vehicles (EVs) has created a growing demand for efficient, sustainable, and convenient charging infrastructure. This paper proposes an IoT-based solar-powered wireless power transfer (WPT) system integrated into roadways for dynamic charging of electric vehicles. The system utilizes photovoltaic panels installed alongside or beneath road surfaces to harness solar energy, which is stored and managed using intelligent energy storage systems. Wireless power transfer is achieved through inductive coupling between transmitter coils embedded in the road and receiver coils mounted on EVs, enabling real-time charging while the vehicle is in motion. The integration of Internet of Things (IoT) technology allows for continuous monitoring, control, and optimization of energy distribution, traffic conditions, and system performance. Sensors and communication modules collect real-time data, which is processed to ensure efficient power delivery and load balancing. This approach minimizes range anxiety, reduces dependency on stationary charging stations, and enhances energy efficiency by utilizing renewable energy sources. The proposed system presents a scalable, eco-friendly, and smart solution for future transportation infrastructure, contributing to reduced carbon emissions and sustainable urban mobility.

DOI: http://doi.org/10.5281/zenodo.379

Impact Of Employee Training On Organizational Productivity

Authors: Dr. Viji R, Director, Prof. (Dr.) Vellayan Srinivasan, Dr.V.O.Kavitha

Abstract: Employee training is a crucial human resource management practice that enhances employee skills, knowledge, and competencies, thereby improving organizational productivity. This study examines the impact of employee training on organizational productivity through an empirical analysis of 100 employees from different departments in selected organizations. Primary data were collected using a structured questionnaire and analyzed using percentage analysis, mean scores, correlation, and regression techniques. The findings reveal that effective employee training significantly improves employee performance, job efficiency, and overall productivity. The study concludes that systematic and continuous training programs are essential for organizational growth and competitiveness.

DOI: https://doi.org/10.5281/zenodo.19590742

Digital Event Managment System

Authors: Prasanna R, Anbarasan M, Nithish Murugesan, Gokulapriyan S, Dr. Rajini S

Abstract: Digital event management faces significant challenges due to the increasing demand for virtual and hybrid events and the lack of integrated platforms for managing them efficiently. Event organizers often rely on multiple disconnected systems for handling registrations, scheduling, live streaming, and participant engagement, which leads to operational complexity and inconsistent user experiences. This project proposes an intelligent and automated digital event management system that uses modern web technologies, cloud services, and real-time communication modules to manage and streamline event operations. The system employs a centralized web-based platform integrated with user interfaces for real-time event coordination using scalable frontend and backend frameworks. A role-based access control mechanism and automated workflow system manage activities such as registration, ticketing, session scheduling, and live interaction. The platform integrates engagement tools including chat, polls, and Q&A features to enhance user participation. This automated approach improves event efficiency, reduces manual coordination effort, minimizes system fragmentation, and supports the growing demand for scalable and user-centric digital event solutions. Digital event management, Web application, Virtual events, Cloud computing, Real-time systems, User engagement, Event analytics, Automated workflow, Scalable platforms, Hybrid events

An Intelligent Machine Learning Framework For Detecting QUIC-Based Traffic Flood Attacks In Encrypted HTTP/3 Networks

Authors: Mr. M. V. Rajesh, Balla Aarathisree, S S V Sumanvitha Palivela, Nagala Bhavya Pragna, Kamalesh Chitra, Taneti Ritesh

Abstract: The rapid growth of encrypted internet protocols such as HTTP/3 and QUIC has significantly improved communication speed and security on modern networks. However, these protocols also introduce new challenges for network security, particularly in detecting Distributed Denial of Service (DDoS) traffic flood attacks. Traditional monitoring techniques rely on packet inspection, which becomes difficult when network traffic is encrypted. This study proposes an intelligent machine learning framework for detecting QUIC-based traffic flood attacks in encrypted HTTP/3 network environments. The proposed system analyses network flow behaviour rather than packet content, enabling effective detection even when traffic payloads are encrypted. To build the detection model, network traffic data are captured and processed into flow-based features such as packet rate, packet size distribution, inter-arrival time, and connection statistics. Data preprocessing techniques are applied to prepare the dataset for machine learning training. Multiple classification algorithms including Logistic Regression, Support Vector Machine (SVM), K-Nearest Neighbours (KNN), and Random Forest are implemented and evaluated using standard performance metrics such as accuracy, precision, recall, F1-score, and ROC–AUC. Experimental results demonstrate that the Random Forest classifier achieves the highest detection accuracy and provides reliable performance for distinguishing between normal and malicious QUIC traffic patterns. To improve transparency and interpretability of the prediction process, Explainable Artificial Intelligence (XAI) techniques such as SHAP and LIME are incorporated into the framework. These methods highlight the most influential network features contributing to attack detection and help security analysts understand the reasoning behind model predictions. The proposed framework enhances the reliability of encrypted traffic monitoring, improves early detection of QUIC traffic flood attacks, and contributes to strengthening the security of next-generation web communication protocols.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.193

AIS-Shield: Self-Supervised Deep Learning For Detecting Dark Vessel Activity Through Intentional AIS Shutdown

Authors: Mrs. N. V. S. Sowjanya, Chavvakula Lasyavalli, Sunkara Vijay Kishore, Kammakatla Shreya, Yerra Sai Rajesh, Chelli Tarun Teja

Abstract: Maritime surveillance plays a crucial role in ensuring the safety, security, and regulation of activities in open sea environments. One of the major challenges faced by maritime authorities is the detection of vessels that intentionally disable their Automatic Identification System (AIS) transponders to conceal illegal activities such as unauthorized fishing, smuggling, or unauthorized entry into restricted maritime zones. AIS messages transmitted by ships are widely used for monitoring vessel trajectories; however, missing AIS signals may occur due to multiple reasons including satellite reception limitations, weather disturbances, or intentional shutdown of AIS devices. Distinguishing between these scenarios becomes difficult when dealing with massive volumes of satellite AIS data. This study proposes an intelligent deep learning framework for detecting intentional AIS shutdown events using self-supervised learning techniques. The proposed approach processes large-scale AIS datasets collected from satellite-based maritime surveillance systems and extracts trajectory-based features such as vessel position, speed, time intervals between messages, and movement patterns. A transformer-based deep learning architecture is used to analyse sequential AIS message data and predict whether a new AIS message is expected within a specific time window. By comparing the predicted results with the actual observations, the system identifies abnormal missing AIS receptions that may indicate intentional signal shutdown. The self-supervised learning approach allows the model to generate pseudo-labels from unlabelled AIS data, eliminating the need for manually labelled datasets. Experimental analysis demonstrates that the proposed framework can process millions of AIS messages in near real-time while achieving high prediction accuracy in detecting abnormal vessel behaviour. The integration of deep learning techniques improves the reliability and scalability of maritime surveillance systems, enabling authorities to identify suspicious vessel activities more efficiently. This framework contributes to enhancing maritime security, improving monitoring capabilities in open sea environments, and supporting timely detection of illegal maritime operations

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.192

PneuXAI-Net: Real-Time Explainable Deep Learning Framework For Multi-Class Pneumonia Detection Using Chest X-Rays

Authors: Mr. K. Srikanth, Beeraka Sharmila, Puppala Madhuri Lakshmi, Darla Ratan Abhishek, Pitchuka Veerababu, Seeram Jaya Venkata Somesh

Abstract: Pneumonia is a significant respiratory disease and one of the top causes of illness and death around the world, especially among children and the elderly. Timely and accurate diagnosis is essential for effective treatment and better patient outcomes. Recent developments in deep learning, particularly Convolutional Neural Networks (CNNs), have shown impressive results in medical image analysis by automatically identifying important patterns in complex image data. This project introduces a real-time pneumonia identification system that combines CNN-based classification with Explainable Artificial Intelligence (XAI) techniques to improve diagnosis accuracy and model clarity. The proposed system processes digitized chest X-ray images through an efficient preprocessing pipeline. This includes noise removal, image normalization, and background feature consideration before sending the images to a trained deep learning model. The ensemble model merges two strong CNN architectures, VGG16 and ResNet50, and uses their complementary feature extraction abilities to boost classification performance. The model classifies Bacterial Pneumonia, Viral Pneumonia, and normal cases, providing clearer clinical insights. Experimental results show high accuracy, strong sensitivity, and real-time inference capability. This allows for pneumonia detection within seconds, which is vital in clinical settings that need quick diagnoses. To tackle the black-box issue of deep learning models, Explainable AI techniques like Grad-CAM++ (Gradient-weighted Class Activation Mapping++) and Score-CAM are used to visualize the key lung areas that affect the model’s predictions. The system also offers confidence scores with visual explanations, enhancing understanding and aiding clinical decision-making. Overall, the proposed CNN and XAI framework offers an efficient, clear, and clinically helpful solution for automated pneumonia detection. The system has strong potential to assist radiologists, boost diagnostic confidence, and contribute to early disease detection and improved patient care.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.191

AI-Driven Dynamic Pricing System For E-Commerce Using Machine Learning And Business Intelligence Analytics

Authors: Mrs. Ch. Veera Gayatri, Palivela Geethasri, Kothapalli Venkannababu, Madhavarapu Chandhra Sekhar Sri Sai, A Lakshmi Chinmayi, Anupoju Sainadh

Abstract: The rapid growth of e-commerce platforms has increased the need for intelligent pricing strategies that can adapt to continuously changing market conditions. Traditional pricing methods used in online retail are often static and rely heavily on historical data analysis, making them inefficient in responding to dynamic market factors such as customer demand, competitor pricing, and seasonal trends. In modern digital marketplaces, businesses generate large volumes of transactional and behavioural data, which creates opportunities for applying machine learning techniques to improve pricing decisions. This study proposes a machine learning-enabled business intelligence framework for dynamic pricing optimization in e-commerce environments. The proposed system integrates data preprocessing, predictive modelling, and business intelligence analytics to support real-time pricing decisions. During the preprocessing stage, historical pricing data, market trends, competitor price information, and customer behavior patterns are collected and processed to improve data quality and consistency. Support Vector Machine (SVM) is employed as the primary machine learning algorithm due to its ability to handle complex and non-linear relationships within large datasets. The business intelligence component of the framework enables efficient data visualization, monitoring, and analysis of market conditions through interactive dashboards and analytical tools. This integration allows businesses to combine predictive insights from machine learning with data-driven business intelligence reports to determine optimal pricing strategies. The proposed system dynamically adjusts product prices by analyzing multiple influencing factors such as demand fluctuations, competitor behavior, and customer purchasing patterns. Experimental evaluation demonstrates that the integration of machine learning and business intelligence significantly improves pricing accuracy, market responsiveness, and decision-making efficiency. By enabling automated and adaptive pricing strategies, the proposed framework helps businesses maximize revenue, enhance competitiveness, and respond effectively to rapidly changing e-commerce environments.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.190

Blockchain-Based Police Complaint Management System For Secure And Transparent FIR Registration

Authors: Mrs. N. Nikhitha, Malla Sudarsan Sai Sunny, Guthula Surya Sindhu, Neelapalli Sri Durga Abhilasha, Govada Pavan Sai, Palisetti Siva Ram

Abstract: The increasing rate of criminal activities and the limitations of existing police complaint systems highlight the need for a more transparent, secure, and efficient method for managing complaints and First Information Reports (FIRs). In many cases, complaints remain unreported or are not officially registered due to procedural delays, corruption, or lack of proper documentation systems. Although online portals such as the Crime and Criminal Tracking Network and Systems (CCTNS) have been introduced, they still operate on centralized architectures that may suffer from issues such as single points of failure, limited transparency, and vulnerability to data tampering. To address these challenges, this study proposes a blockchain-based police complaint management system designed to provide secure, decentralized, and tamper-proof storage of complaint records. In the proposed system, complaint details and FIR records are encrypted and stored using the InterPlanetary File System (IPFS), while the corresponding hash values are recorded on a blockchain network to ensure immutability and data integrity. The decentralized nature of blockchain technology ensures that complaint records cannot be altered or deleted without network consensus, thereby preventing unauthorized modifications and enhancing system transparency. Additionally, timestamped blockchain entries provide verifiable proof of complaint submission, enabling citizens to demonstrate that their complaint was officially recorded even if authorities deny receiving it. By integrating blockchain with distributed file storage technologies, the proposed system enhances trust between citizens and law enforcement agencies while ensuring secure and transparent management of police complaints. The framework also supports the broader goals of e-governance by improving accountability, data security, and accessibility in public service systems.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.189

Socio Net: An Interpretable Deep Neural Network Framework For Crime Detection In Social Media Platforms

Authors: Mr. V. Hemanth Sai, Devulapalli Srujan, Mattaparthi Teja Nirgun, Mummidi Lohith Naga Ratan, Poluparthi Abhishek, Kasireddi Naga Venkata Sai Navadeep

Abstract: Social media platforms (SMPs) are widely used for communication and information sharing, but they are also increasingly exploited for criminal activities. These activities include forming illegal groups, spreading false information, stealing personal data, and conducting cyberattacks. The ease of access and anonymity provided by SMPs make them attractive for criminals to perform such actions. Sensitive information such as passwords, financial details, and personal data can be misused, leading to serious threats like identity theft, data breaches, and malware attacks. This paper focuses on detecting criminal activities on social media using machine learning techniques. By analyzing user-generated content, the system can identify suspicious patterns and classify potentially harmful activities. The proposed approach aims to improve early detection and help in preventing cybercrimes effectively. Additionally, it highlights the importance of user awareness and responsible data sharing to reduce risks associated with social media usage.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.188

Smart Vision: AI-Powered Traffic Violation Detection Using YOLOv7

Authors: Mrs. K. Tulya Sree Simla, Penmatsa Dhathri Vidya Prabha, Bikkina Anusha, Gubbala Chandra Mouli, Nimmagadda Sanjith, Vakadi Ayyappa Surya Sri Harsha

Abstract: Traffic violations are a major contributor to road accidents and fatalities, especially in densely populated urban regions. Common violations such as jumping red lights, triple riding on two-wheelers, reckless driving, and riding without helmets significantly increase the likelihood of accidents. Conventional traffic monitoring systems largely depend on manual supervision by traffic police or basic sensor-based methods, which are often inefficient, time-consuming, and susceptible to human error. To overcome these limitations, intelligent traffic monitoring systems based on computer vision and deep learning have gained increasing importance. This paper presents a deep learning-based automated traffic violation detection system using the YOLOv7 object detection model. The proposed system processes video streams captured from roadside surveillance cameras and analyses them frame by frame to detect various traffic violations. The YOLOv7 model is used to identify vehicles and generate bounding boxes around detected objects. A predefined threshold line is applied to determine whether a vehicle crosses the signal during a red light, thereby detecting signal violations. Additionally, the system identifies overloading or triple riding on two-wheelers by analysing the number of riders within a single vehicle bounding box. Helmet violations are also detected by determining whether riders on motorcycles are wearing helmets. If a rider is identified without a helmet, the system classifies it as a violation. The system utilizes publicly available datasets such as the MS COCO dataset for vehicle detection and a custom annotated dataset for detecting overloading and helmet violations. The model is trained and evaluated using performance metrics including precision, recall, F-measure, and mean Average Precision (mAP). Experimental results indicate that the proposed system can accurately detect multiple traffic violations while maintaining efficient real-time performance. The proposed approach offers a cost-effective, automated, and scalable solution for traffic monitoring. It can assist traffic authorities in improving road safety and reducing the burden of manual monitoring. Furthermore, the system can be integrated with existing smart city surveillance infrastructure to support intelligent transportation management and law enforcement.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.187

Crop Sense AI: Data-Driven Crop Recommendation Using ML And Deep Learning

Authors: Mrs. K. Harika, Rowthu Kavyanjali Priya, Madeti Vineetha, Samanthakurthy Rajavardhan, Sachin Pandit, Penky Adi Seshu

Abstract: Agriculture plays a vital role in maintaining food security and contributing to the global economy. However, selecting the most appropriate crop for a specific region remains a significant challenge for farmers due to variations in soil nutrients, climatic conditions, and environmental factors. Incorrect crop selection can result in low productivity, inefficient resource utilization, and financial losses. With the growing availability of agricultural data and advancements in artificial intelligence, machine learning techniques have become effective tools for improving decision-making in agriculture. This study proposes an intelligent crop recommendation system that combines machine learning and deep learning models to assist farmers in selecting the most suitable crop based on soil and environmental conditions. The system analyses key agricultural parameters such as nitrogen (N), phosphorus (P), potassium (K), rainfall, soil pH, temperature, and humidity. These features are used to train predictive models capable of recommending the most appropriate crop for cultivation. Various machine learning and deep learning algorithms, including Decision Tree, Random Forest, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Naïve Bayes, Artificial Neural Networks (ANN), Deep Neural Networks (DNN), and Temporal Convolutional Networks (TCN), are implemented and evaluated. The models are trained using a publicly available agricultural dataset containing multiple crop types along with environmental attributes. Performance is assessed using evaluation metrics such as accuracy, precision, recall, and F1-score to identify the most effective model. Experimental results show that ensemble and deep learning models achieve high prediction accuracy in recommending suitable crops. The system also provides a user-friendly interface that enables farmers to input soil and environmental parameters and receive crop recommendations in real time. The proposed approach supports precision agriculture by enabling data-driven farming practices, improving crop yield, and assisting farmers in making informed decisions.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.186

An AI-Driven Real-Time Parking Monitoring And License Plate Recognition System Using CCTV

Authors: Mrs. A. Daiva Krupa Nirmala, Polina Sai Satvika, Vennapu Lingeswara Rao, Dasari Manvanth, Syed Irfan

Abstract: This paper proposes a smart car parking system that uses image processing and real-time CCTV monitoring to efficiently detect parking space availability and recognize vehicle license plates. The system is designed for both open parking areas and multi-storey parking environments. It uses Python along with the OpenCV library to analyze video input and determine whether parking slots are occupied or vacant based on pixel-level analysis and image processing techniques. In addition, the system integrates Optical Character Recognition (OCR) using the Tesseract engine to automatically extract license plate information from captured images. To improve accuracy, multiple preprocessing techniques are applied to handle variations in image quality, lighting, and noise. The proposed system enables automated parking management, reduces manual effort, and enhances monitoring efficiency, making it suitable for real-time smart parking applications.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.185

A YOLOv5-Based Framework For Real-Time Wildlife Detection And Intrusion Alert Systems

Authors: Mrs. D. Chakra Satya Tulasi, Bejjipuram Jahnavi, Yarlapati Venkata Naga Durga Varun, Guraja Jayachandra, Peruri Vinay

Abstract: An advanced wild animal detection and alert system is developed using the YOLOv5 (You Only Look Once version 5) model. The system uses an object detection algorithm to identify wild animals and provide real-time alerts to users. A camera captures live video footage, which is processed by a computer running the YOLOv5 model to accurately detect and classify animals. When a wild animal is detected, the system immediately generates alerts such as warning sounds or notifications to prevent potential danger. These alerts can also act as deterrents to scare animals away and improve safety. The system is useful in areas where wild animal movement is common, such as forest borders, agricultural fields, and highways. Overall, the system provides an efficient and real-time solution for monitoring wildlife and reducing human-animal conflicts. Future improvements can focus on increasing accuracy and enhancing real-world performance under different environmental conditions.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.184

Life Sense: A Deep Learning-Based Framework For Mechanical Components Health Monitoring And Life Prediction

Authors: Mrs. V. Suvarna, Mavuri Bhuvana, Boddu Rajeev, Choppella Vamsi Kumar, Dhulipudi Sree Vivek

Abstract: To improve prediction accuracy and enable real-time monitoring of mechanical components, a deep learning-based approach is proposed. The system utilizes a Convolutional Neural Network (CNN) to extract important features from mechanical parts using sensor data. These features are further processed through fully connected layers for information fusion and classification, allowing accurate prediction of remaining useful life and health status. The trained deep learning model is integrated into a monitoring system to create a complete framework for continuous condition monitoring and life prediction. The system is further optimized to enhance prediction accuracy, real-time performance, and adaptability under different working and environmental conditions. Experimental results show that the proposed model achieves high performance with a Mean Absolute Error (MAE) of 2.1, Root Mean Squared Error (RMSE) of 2.5, and Mean Absolute Percentage Error (MAPE) of 10%. These results demonstrate the effectiveness of the approach and its potential for practical applications in industrial maintenance and reliability improvement.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.183

A Context-Aware Multimodal Explainable Deep Learning Framework For Robust Android Malware Detection And Proactive Threat Prevention

Authors: Mr .B.Janu Naik, Velugubanti Lakshmana Siva Ganesh, Vignesh Mullangi, Vanapalli Veera Satya Sai Praneesh, Maddula Veera Venkata Sai Pradeep

Abstract: With the rapid expansion of Android applications, malware attacks targeting mobile devices have increased significantly, creating serious security and privacy concerns for users. Traditional malware detection approaches, such as signature-based and rule-based methods, often fail to detect newly emerging or obfuscated malware variants. To overcome these challenges, this study proposes an explainable artificial intelligence-based framework, named XAI-Droid, for effective Android malware detection and classification. The proposed system integrates deep learning techniques with explainable AI (XAI) methods to enhance detection accuracy while ensuring transparency and interpretability in decision-making. Feature extraction is carried out using static analysis techniques, and the extracted features are used to train advanced machine learning and deep learning models. To improve trust and reliability, explanation methods such as feature importance analysis are incorporated to identify the key attributes influencing classification outcomes. Experimental results demonstrate that the proposed framework achieves high detection accuracy while maintaining interpretability, making it suitable for practical cybersecurity applications. By combining strong classification performance with explainability, XAI-Droid contributes to the development of reliable and trustworthy AI-based mobile security systems.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.182

A Machine Learning-Based Automated System For Early Detection And Classification Of Hearing Loss In Infants And Toddlers

Authors: Ms. Y Suma Chamundeswari, Nalli Neeharika, Sneha Dindi, Abbireddy Durga Devi, Nyasavarajula R S Gowtham Datta, Ayanamahanthi Thandava Krishna Murthy

Abstract: Hearing impairment is one of the most common sensory disorders affecting newborns, infants, and young children worldwide. Early detection of hearing loss is crucial because delayed diagnosis can negatively affect speech development, cognitive growth, social interaction, and educational outcomes in children. However, many developing and underdeveloped regions face a shortage of audiologists and otolaryngologists, which often results in delayed diagnosis and limited access to hearing care services. This situation highlights the need for automated and intelligent diagnostic tools that can assist healthcare professionals in identifying hearing impairments more efficiently. This study proposes an automated hearing loss detection framework based on machine learning techniques to support medical professionals in diagnosing hearing impairments in newborns, infants, and toddlers. The proposed system integrates a hearing test data generation module with a machine learning classification model capable of analyzing audiometry test data and predicting the presence and characteristics of hearing loss. The data generation module creates a comprehensive dataset representing different hearing conditions, which is then used to train and evaluate the machine learning model. By employing multiclass and multi-label classification techniques, the model can identify the type, degree, and configuration of hearing loss with high accuracy. Experimental results demonstrate strong diagnostic performance, achieving a prediction time of approximately 634 milliseconds, a log-loss reduction rate of 98.48%, and macro and micro precision values close to 100%. These results indicate that the proposed framework can provide rapid and reliable diagnostic support for healthcare professionals, enabling earlier intervention and improving access to hearing care in regions with limited medical resources.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.181

An Intelligent Time Series Forecasting Model For Financial Market Prediction Using Support Vector Machine

Authors: Mr.B. Janu Naik, Hemadri Sumedha, Anala Sanghavi, Giduthuri Charanchandu, Sathi Sanjana Reddy, Geetha Yeswanth Kumar

Abstract: Forecasting financial market trends has become an important task for investors, financial institutions, and analysts due to the increasing complexity and volatility of modern financial systems. Accurate prediction of market movements such as stock prices, exchange rates, and commodity prices can significantly assist in making informed investment decisions and managing financial risks. However, financial market data is highly dynamic, nonlinear, and influenced by multiple economic and external factors, which makes accurate forecasting a challenging problem for traditional statistical methods. In this study, a machine learning-based framework is proposed for forecasting financial market trends using time series analysis. The proposed approach utilizes historical financial data including stock prices, trading volumes, and other relevant financial indicators to train predictive models capable of identifying patterns and relationships within the data. A Support Vector Machine (SVM) algorithm is employed as the primary forecasting model due to its strong capability to handle nonlinear relationships and high-dimensional datasets effectively. The system performs several important stages including data loading, preprocessing, feature selection, model training, and performance evaluation. During preprocessing, missing values and irregularities in the dataset are handled, and normalization techniques are applied to ensure consistent feature scales. The processed data is then used to train the SVM model, which learns complex patterns present in historical financial data to generate predictions for future market trends. The performance of the forecasting model is evaluated using standard evaluation metrics such as Mean Squared Error (MSE) and Mean Absolute Error (MAE) to measure prediction accuracy. Experimental analysis demonstrates that the proposed machine learning framework is capable of effectively capturing nonlinear patterns present in financial time series data and providing reliable forecasting results. By leveraging machine learning techniques, the proposed system improves prediction efficiency and supports intelligent decision-making for traders, investors, and financial analysts. The framework can serve as a valuable tool for financial market analysis and can be further enhanced by integrating hybrid machine learning models and real-time financial data sources.

DOI: http://doi.org/10.61137/ijsret.vol.12.issue2.180

ChatGMVIT: An AI-Powered Academic Assistance Chatbot Using Firebase And Gemini AI

Authors: Chetan Uddhav Wagh, Yash Tatyaba Khandare, Shreya Prashant Mhaskar, Prof. K. R. Metha

Abstract: Artificial Intelligence based conversational systems are transforming how information services are delivered in many fields including education. Students frequently require quick access to academic information such as course details, admission procedures, faculty information, and examination schedules. However, traditional communication channels such as notice boards, websites, and help desks often fail to provide instant responses. This research proposes ChatGMVIT, an AI-powered chatbot designed to assist students by providing instant responses to institutional queries. The system is implemented as an Android application integrated with Firebase Firestore as a knowledge database and the Gemini 2.5 Flash Lite model for natural language processing.

DOI: https://doi.org/10.5281/zenodo.19593570

Hybrid AI-Agent Driven Process Optimization Framework For Enterprise Decision System

Authors: Dr. Nilesh Jain, Kishan Vyas, Monil Lalwani, Kushal Shah

Abstract: The scope of enterprise reporting systems has grown significantly as far as data size, dashboard interactiveness, and visual analytics are concerned. However, in most companies, these systems operate as a static support for making decisions instead of an active decision system. The issue is that even though the dashboards detect patterns, anomalies, and trends, the user should understand these results and trigger the next steps. This implies some time lag, inconsistencies, and dependency on human decision-making. Thus, in this article, we propose a mixed AI-agent framework of enterprise decision systems where the intelligence ability of artificial intelligence is complemented by the execution capabilities of intelligent agents through staged development. The framework involves data collecting, cleaning, transformation, creation of dashboards, analysis with the help of machine learning algorithms for predictions, classifications, clustering, and detecting anomalies, and finally, the use of agents that translate model outputs into workflow activities such as approval, escalating, prioritizing, and notifying. The primary idea of this research is to split the process of intelligence generation and actions taking at different process stages, thereby increasing the modularity, interpretability, and effectiveness. However, it retains its practical applicability and relevance of dashboard analysis. The paper discusses the conceptual framework, technical design, model types, logic of process mapping, criteria of assessment, governance issues, and the possibility of enterprise implementation of the proposed approach

Alumni Connect: Enhancing Alumni Networking And Support With Ai Assistance Using Langgraph And Pinecone

Authors: Nitish Prajapati, Mahfooz Khan, Piyush Bagadi, Suraj Kumar

Abstract: In order to build solid relationships between colleges, their alumni, and present students, alumni en- gagement is essential. A specialized platform called ALUMNI CONNECT: BRIDGING THE GENERATION was created to strengthen these connections by offering a controlled, engaging envi- ronment for networking, career development, and mentoring. Key elements of the platform include job advertising, mentorship programs, discussion boards, guidance requests, regional alumni filters, alumni contributions, including donations, lab installations, and event planning. Furthermore, it maintains a trustworthy community by removing phony accounts and verifying profiles to guarantee authentic- ity. The initiative’s main driving forces are the underutilization of alumni networks and the dearth of easily accessible, individualized student counseling. By tackling these issues, the platform encour- ages deep connections, allowing alumni to commemorate successes and offer insightful career guidance and university-specific recommendations. Students can close the gap between academic accomplish- ment and career success by gaining access to industry connections and mentorship possibilities through increased engagement. In order to improve professional development and institutional linkages,this project seeks to establish a smooth and effective alumni-student engagement paradigm. ALUMNI CONNECT hopes to create a thriving self-sustaining ecosystem that benefits both students and alumni by utilizing technology to promote involvement.

AI-Driven Augmented Reality Based Smart Campus Navigation System

Authors: Dr. S. Sharon Priya, Anish Fathima N, Blessy Evangeline A

Abstract: The campus environment in universities includes many blocks laboratories offices hostels libraries and other facilities spread across different areas which makes navigation difficult for students visitors and staff. The existing methods such as signboards printed maps and verbal directions do not provide proper support and users often struggle to find the correct location. The absence of real time guidance creates confusion especially during busy periods like admissions examinations and events where many people move at the same time. The difficulty increases when users are not familiar with the campus layout and this leads to time loss and frustration. The need for a simple reliable and easily accessible navigation system becomes very important in such situations. The proposed system is an AR based smart campus navigation solution that works through a mobile browser without the need for any application installation. The system uses GPS and device sensors to continuously track user position and direction and provides accurate navigation support. The A star algorithm is used to calculate the shortest path between the current location and the destination and this improves navigation efficiency. The guidance is shown using augmented reality arrows on the camera view which makes directions easy to understand and follow. The system also includes an AI chatbot for user queries voice guidance for hands free navigation and crowd prediction to avoid busy areas. The Firebase database is used to provide real time information such as lab availability staff details and event updates. The system provides a complete and user friendly solution for campus navigation.

Schedulify: A Hybrid Approach for Automated University Timetable Generation

Authors: Biju Balakrishnan, Abdul Aziz Khatri, Hemang Korane, Yeshang Upadhyay, Eyakramul Hussain, Akshay Nugurwar

Abstract: This project provides a hybrid solution for the automated generation of timetables through the combination of Linear Programming (LP) and Constraint Satisfaction Problem (CSP) solutions. It’s intended to address the NP-hard problem of university course timetabling, which involves the satisfaction of a considerable number of conflicting constraints and preferences. The method for solving this problem involves two distinct steps. In the first step, a linear programming method is employed to determine the optimal assignment of teachers to subjects. This is a global optimisation technique that aims to maximize the satisfaction of teacher preferences and ensure a well-balanced allocation of teaching loads for all faculty members. In the second phase, a CSP algorithm with backtracking is applied to these optimal assignments to further assign time slots and classrooms. This phase is handled by both hard and soft constraints. Hard constraints, such as the unavailability of a teacher or a classroom, are represented by hard constraints that should not be violated to avoid any scheduling conflicts. Soft constraints, such as teacher time slot preferences and constraints on the lecture interval, are employed to filter the most optimal assignments from the pool of feasible solutions to enhance the quality of the schedule.

DOI: https://doi.org/10.5281/zenodo.19594429

Real-Time Traffic Sign Recognition Using YOLOv7: A Robust Deep Learning Approach for Autonomous Driving

Authors: Parag Hossain

Abstract: Traffic Sign Recognition (TSR) is a critical component of autonomous driving systems and Advanced Driver Assistance Systems (ADAS). However, real-world environmental challenges such as occlusions, lighting variations, multi-scale changes, and motion blur often degrade the performance of traditional vision pipelines. This paper presents a robust real-time TSR system based on the YOLOv7 architecture. The proposed model leverages an E-ELAN backbone for hierarchical feature extraction, a PANet neck for multi-scale semantic fusion, and an anchor-free IDetect head for precise localization. Trained on a custom traffic_sign_data dataset with aggressive data augmentation, the system achieves 94% mAP@0.5 and 38% mAP@0.5:0.95 at 45 frames per second on an NVIDIA RTX 3090 GPU. Comparative evaluations show that YOLOv7 significantly outperforms YOLOv5 with 89% mAP@0.5 and Faster R-CNN with 91% mAP@0.5 at only 12 FPS. The model is further optimized via ONNX-to-TensorRT conversion, enabling efficient deployment on edge computing platforms such as NVIDIA Jetson AGX Xavier.

DOI: https://doi.org/10.5281/zenodo.19594864

Real-Time Vehicle Detection, Tracking And Recognition Using YOLOv26 (Ultralytics)

Authors: Atchaya K, Madhumitha T, Pasumarthini R, Siva Sandhiya M, Susmitha S

Abstract: The rapid growth of urbanization has resulted in increased vehicle density on roads, raising the demand for efficient and intelligent traffic monitoring systems. This paper presents a real-time vehicle detection, tracking, and recognition system using YOLOv26(ultralytics), the latest advancement in the You Only Look Once (YOLO) architecture. The proposed system leverages deep learning-based object detection to detect and classify vehicles from video streams captured by surveillance cameras. YOLOv26(ultralytics) offers improved accuracy and speed over its predecessors, making it highly suitable for real-time Intelligent Transportation System (ITS) applications. The system incorporates Deep SORT for robust multi-object tracking and supports recognition based on vehicle attributes including color, type, and license plate.

ADAP (Automated Data Analytics Platform): A Data Intelligence Pipeline With Expert Verification For Enterprise-Grade AI-Driven Data Quality, Validation, And Adaptive Analytics

Authors: Aayush Yogesh Sanklecha, Pranav Dattatray Gund, Aditya Yogesh Salunke, Samarth Pramod Koli, Prof Mr.H.B.Gadekar

Abstract: Data quality remains the critical bottleneck in enterprise machine learning pipelines. Unreliable, schema-broken, drifted, or regulatory non-compliant data causes downstream analytics failures with consequences ranging from inaccurate predictions to regulatory penalties. This paper presents ADAP (Automated Data Analytics Platform) (Data Intelligence Pipeline with Expert Verification), an end-to-end data intelligence platform that unifies multi-source data ingestion, NLP-augmented semantic schema classification, seven-dimensional parallel validation, regulatory compliance enforcement (AML, HIPAA, SOX, GDPR), AutoML with SHAP explainability, and a dual reinforcement learning (RL) adaptation engine — all within a single auditable medallion-architected system. ADAP (Automated Data Analytics Platform) achieves schema classification accuracy of 94.7% across 31 semantic types, anomaly detection AUROC of 0.961, calibrated confidence scoring with ECE 0.0225 and AUC 0.9784, and multivariate drift detection at 89.4% accuracy at moderate distributional shift. A PPO Actor-Critic agent pre-trained over 1,000 synthetic episodes and warm-started via Thompson Sampling adapts 8-axis pipeline execution strategies in real time. End-to-end pipeline latency is under 7.4 seconds for 100,000-row datasets. All six production models pass tightened v7 quality gates, with performance validated end-to-end on held-out enterprise datasets.

CodeFox: A Modular Platform For Repository Insights

Authors: Utsav R. Hirapra

Abstract: As the scope of software projects increases, it becomes increasingly difficult to ensure proper code quality and conduct efficient code reviews, not because the required tools lack, but because all relevant information tends to become buried under unnecessary noise. In an attempt to combat that issue, CodeFox is presented as a lightweight, modular solution for discovering valuable insights by leveraging modern AI algorithms. In terms of functionality, CodeFox consists of a developer-friendly user interface, various automated review tools, and a meaning- based code search module. The platform collects metadata such as commits, reviews, comments, as well as ownership information and uses semantic embeddings to index critical elements within the code. By leveraging vector search algorithm, CodeFox allows its users to easily discover similar code, discussion threads related to the code, as well as reviewers who were working on this code. The use of the platform at an early stage of development in our company allowed us to shorten the review cycle process as well as reveal the reasoning behind code changes more efficiently. In this paper, we discuss CodeFox architecture, key aspects of integration, namely webhooks, background workers, and persistent storage, as well as share our experience of implementing a convenient yet lightweight platform to facilitate further development. We plan to conduct a user study in order to evaluate efficiency in the context of the problem and implement more advanced data gathering features and intelligent suggestions in the future.

DOI: https://doi.org/10.5281/zenodo.19603725

The Study of Goods and Services Tax Implementation

Authors: Mr. Deva Dorin DR, Associate Professor Dr. T. M. Hemalatha

Abstract: The Goods and Services Tax (GST) is one of the most significant tax reforms introduced in India to simplify the indirect taxation system. It was implemented on 1 July 2017 with the objective of replacing multiple indirect taxes such as Value Added Tax (VAT), Service Tax, Excise Duty, and others into a single unified tax structure. The main purpose of GST is to create a common national market, reduce tax cascading, and improve transparency in the taxation system. This study focuses on the implementation of GST and its impact on businesses, consumers, and the overall economy. The research examines the advantages, challenges, and effectiveness of GST after its implementation. It also analyzes how GST has simplified tax compliance, improved tax collection, and influenced the pricing of goods and services.The study is based on secondary data collected from journals, government reports, articles, and online sources. The findings indicate that GST has brought significant changes to the Indian tax structure by promoting transparency and reducing tax complexities, although certain challenges such as compliance issues and technical difficulties were faced during the initial stages of implementation. Overall, the implementation of GST has contributed to strengthening India’s taxation framework and supporting economic growth by creating a more efficient and unified tax system.

DOI: https://zenodo.org/records/19605629

A Study on the Role of Microfinance in Promoting Entrepreneurship Among Women

Authors: Mr. Albino Albert Raj Jenobe, Associate Professor Dr. T. M. Hemalatha

Abstract: Microfinance has become a significant financial instrument for fostering entrepreneurship within economically disadvantaged communities, especially among women. Women frequently encounter obstacles such as insufficient access to credit, limited financial knowledge, and societal constraints that hinder their full engagement in entrepreneurial endeavors. Microfinance institutions (MFIs) tackle these issues by offering small loans, savings options, insurance, and financial assistance to those who are marginalized by conventional banking systems. This study examines the role of microfinance in promoting entrepreneurship among women and its impact on economic empowerment and social status. The research adopts a descriptive research design and collects data through structured questionnaires from women beneficiaries of microfinance services. Secondary data sources such as journals, reports, and research publications are also used to support the study. The findings reveal that microfinance services have significantly contributed to the development of women entrepreneurs by enabling them to establish small businesses, improve income levels, and achieve financial independence. The study also highlights the challenges faced by women entrepreneurs and suggests measures to strengthen microfinance programs for sustainable economic development.

DOI: https://zenodo.org/records/19605728

Smart-Kheti: An AI-Powered Smart Agriculture Platform For Crop Recommendation, Disease Detection, And Yield Prediction

Authors: Rohit singh, Vikas pal, Minal suthar, Priti tangadi

Abstract: Agriculture forms the backbone of the Indian economy, yet smallholder farmers continue to face critical challenges including crop failure, rampant plant disease, unpredictable weather, and limited access to expert advisory services. This paper presents Smart-Kheti, a web-based AI-powered smart agriculture platform designed to democratize data-driven decision support for farmers. The proposed system integrates a personalized crop recommendation engine utilizing soil nutrient parameters (N, P, K), pH, temperature, humidity, and rainfall processed through an XGBoost-based multi-class classifier; an automated plant disease detection module employing a Convolutional Neural Network (CNN) trained on the PlantVillage dataset and deployed via TensorFlow Lite for server-side inference and TensorFlow.js for offline client-side inference; and a yield prediction module utilizing XGBoost regression on multi-year historical agricultural data. The platform employs a full- stack architecture with React.js and TypeScript on the frontend and Python FastAPI on the backend, containerized using Docker for scalable deployment. Additional features include a profit calculator, real-time market insights from government data APIs, offline support, and multilingual accessibility. Experimental evaluation demonstrates crop recommendation accuracy of 97.4%, disease detection accuracy of 93.7%, and yield prediction RZ of 0.87.

Artificial Intelligence In Business Decision Making; Opportunities And Challenges

Authors: Mohamed Issam S, Associate Professor Dr. T. M. Hemalatha

Abstract: Artificial Intelligence (AI) is a key driver of modern business transformation, especially in corporate environments where data-driven decision-making remains complex. AI-driven tools play a significant role in bridging this gap by providing scalable predictive analytics, automated reporting, and risk assessment to businesses. This study examines the role of Artificial Intelligence in enhancing business decision-making. The research focuses on understanding the accessibility, efficiency, and effectiveness of AI services in improving the operational and strategic outcomes of businesses. Using a descriptive research design, data was collected through a structured questionnaire and supported by secondary sources. The findings indicate that AI has contributed positively to decision-making by enhancing data processing speed, encouraging proactive strategies, supporting revenue-generating activities, and reducing dependence on manual heuristics. The study highlights the importance of strengthening AI integration practices to achieve sustainable business growth.

DOI: https://zenodo.org/records/19605830

A Study of Social Media Marketing in Shaping E-Commerce Success

Authors: A Banu, Associate Professor Dr. T. M. Hemalatha

Abstract: Social media marketing has emerged as a powerful tool in influencing consumer behavior and driving the success of e-commerce businesses. With the rapid growth of digital platforms such as Instagram, Facebook, YouTube, and WhatsApp, online retailers increasingly rely on social media to attract, engage, and retain customers. This study aims to analyze the role of social media marketing in shaping e-commerce success by examining consumer awareness, purchasing behavior, engagement levels, and perceived effectiveness of social media campaigns. Primary data were collected through a structured questionnaire from 200 respondents actively involved in online shopping. Statistical tools such as Percentage Analysis, Correlation, Chi-Square Test, and One-Way ANOVA were applied to analyze the data. The findings reveal a significant relationship between social media marketing strategies and e-commerce performance, indicating that social media plays a crucial role in enhancing brand visibility, customer trust, and sales growth.

DOI: https://zenodo.org/records/19605881

The Future Of Authentication With FIDO: Beyond The Binary Assertion

Authors: Kritika Kumari Ojha

Abstract: Phishing remains a critical cybersecurity threat, as traditional blacklist-based systems struggle against rapidly evolving domains and zero-day attacks. While machine learning (ML) has emerged as an adaptive solution for detection, the ultimate defense lies in re-engineering the authentication handshake itself. This paper explores the transition from “pass/fail” binary assertions toward a richer, contextual verification ecosystem powered by FIDO (Fast Identity Online) standards. We analyze how WebAuthn and CTAP2 shift the paradigm from possession-based secrets to high-assurance, phishing-resistant identity verification.

DOI:

A Study on the Impact of Artificial Intelligence in Transforming Modern Business and Strategies for Adaptation

Authors: Mr Gipson M, Assistant Professor Ms. Bushra B

Abstract: Artificial Intelligence (AI) has emerged as a transformative force reshaping modern businesses across industries. It enables automation, enhances decision-making, and improves customer experience. This study examines how AI is transforming business operations, marketing strategies, and organizational efficiency. A descriptive research design was used, and data was collected through structured questionnaires and supported by secondary data sources. The findings reveal that AI significantly increases productivity, reduces operational costs, and enhances customer satisfaction. Businesses that adopt AI gain a competitive advantage in the market. However, challenges such as high implementation cost, lack of skilled workforce, and ethical concerns still exist. The study concludes that effective AI adoption strategies are essential for sustainable business growth.

DOI: https://zenodo.org/records/19606178

Classification And Performance Analysis Of Power Management Strategies In Wireless Sensor Networks

Authors: Priya Yadav

Abstract: Wireless Sensor Networks (WSNs) have become an essential technology for monitoring and data collection in various applications such as environmental monitoring, smart cities, healthcare systems, industrial automation, and military surveillance. These networks consist of numerous sensor nodes that are deployed in remote locations and operate with limited battery power. Since replacing or recharging batteries is often difficult, efficient power management becomes a crucial factor in ensuring the long-term operation of wireless sensor networks. Power management strategies aim to reduce energy consumption while maintaining network performance and reliability. This review paper presents a classification of major power management strategies used in wireless sensor networks and analyzes their performance based on key parameters such as energy consumption, network lifetime, scalability, and reliability. The study categorizes power management techniques into duty-cycling mechanisms, transmission power control methods, energy harvesting techniques, and energy-efficient routing approaches. Furthermore, recent advancements such as artificial intelligence-based energy optimization and adaptive power management techniques are discussed. The paper provides insights into current research trends and identifies future research directions for improving energy efficiency in wireless sensor networks.

</p

Machine Learning For Image Restoration: A Review Of Methods, Trends, And Challenges

Authors: Shradha Kumavat, Kapil Shah

Abstract: Image restoration the task of recovering degraded or damaged images has become essential across many technical domains, including space imaging, medical imaging, and several post-processing applications. Most restoration techniques begin by modeling the degradation process that corrupts an image, typically involving blur and noise, and then attempt to reconstruct an approximation of the original image. However, in real-world scenarios, degradation is often unknown, requiring the simultaneous estimation of both the true image and the blurring function directly from the observed degraded image, without relying on prior knowledge of the blur mechanism. This thesis proposes a novel digital image restoration approach based on punctual kriging, supported by multiple machine learning algorithms. The work focuses on restoring images corrupted by Gaussian noise by achieving an effective trade-off between two competing goals: producing smooth, visually pleasing results while preserving edge details and structural integrity.

Automating The Penetration Testing Flow Using Model Context Protocol (MCP) And AI-Orchestrated Security Agents

Authors: Navipriyaa M, Pooja Ponrani D, Prathiba Devi V S, Mr. Prasannavenkatesan K

Abstract: Penetration testing plays a vital role in identifying security weaknesses in modern computing systems. With the rapid growth of distributed architectures, cloud-native applications, and microservices, traditional penetration testing approaches have become increasingly complex and time-consuming. Although automated tools are widely used, they typically function in isolation and require significant human expertise to coordinate multi-stage attack scenarios This paper presents an enhanced AI-driven penetration testing framework that leverages the Model Context Protocol (MCP) for structured communication and orchestration among multiple intelligent agents. The proposed system integrates reconnaissance, vulnerability assessment, exploitation, privilege escalation, and reporting into a cohesive pipeline. Unlike traditional systems, the framework incorporates contextual reasoning, adaptive decision-making, and dynamic exploit chaining using an AI Planner. Additionally, the system constructs real-time attack graphs and computes risk scores based on vulnerability severity, exploit confidence, and attack depth. Experimental results demonstrate significant improvements in automation efficiency, reduction in manual effort, and higher success rates in identifying complex exploit chains. The proposed framework represents a shift from static automation toward intelligent, adaptive penetration testing systems.

Hybrid Deep Learning-Based Artificial Intelligence Framework For Early Cancer Detection And Preventive E-Healthcare Systems

Authors: Ms. Babita, Dr. Brij Mohan Goel

Abstract: Cancer will continue to be a leading cause of mortality worldwide, making early detection and timely intervention essential for improving survival rates. This study will propose a hybrid Artificial Intelligence (AI)-based healthcare framework for early cancer detection and preventive analysis using deep learning techniques. The model will integrate Convolutional Neural Networks (CNN) for medical image feature extraction and Long Short-Term Memory (LSTM) networks for analyzing sequential clinical data.The system will be evaluated on benchmark cancer datasets using performance metrics such as accuracy, precision, recall, and F1-score. The proposed hybrid model is expected to outperform traditional machine learning approaches by achieving higher accuracy and lower error rates.The framework will support early-stage diagnosis, risk prediction, and personalized preventive strategies. Although challenges such as computational complexity and data privacy will persist, the proposed system is anticipated to offer strong potential for real-world healthcare applications and contribute to AI-driven cancer care.

Entropic-Topological Barycentric Synthesis For GNSS RTK Averaging

Authors: Sandeep Kumar Kashyap, Shweta Vikram

Abstract: High-precision GNSS Real-Time Kinematic (RTK) positioning often suffers from gross errors caused by non-line-of- sight (NLOS) multipath and other anomalies, which can dramatically bias simple coordinate averages. This paper presents Entropic-Topological Barycentric Synthesis (ETBS), a novel framework that dynamically selects a reliable subset of GNSS coordinates and computes a weighted barycentric average. The method proceeds in phases: (1) Topological filtering of the raw point set using kernel density estimation to identify and remove outliers; (2) Entropy weighting of remaining points based on multiple quality metrics (e.g. carrier-to-noise ratio, PDOP, satellite elevation variability) to assign higher weight to more reliable observations; and (3) Barycentric coordinate synthesis by computing the Wasserstein (transport) barycenter of the weighted points, yielding the final coordinate estimate. In synthetic tests mimicking open-sky and harsh urban conditions, ETBS consistently isolates outliers and yields centimeter-level accuracy, whereas traditional mean/median or robust least-squares methods produce errors on the order of decimeters or more. The results demonstrate that ETBS effectively neutralizes extreme outliers and achieves superior positioning precision.

DOI: https://doi.org/10.5281/zenodo.19606990

Ecological Significance Of Ruminant Microbial Symbiosis: Nutrient Cycling, Climate Impact, And Sustainable Agriculture

Authors: Dr. Jyoti Prakash

Abstract: Ruminant animals have a highly specialized microbial ecosystem within their rumen, allowing for the digestion of complex plant material such as cellulose. This mutualistic relationship not only provides for the nutritional requirements of the host animal but is also essential for ecosystem functioning. The rumen microbes play a large part in carbon and nitrogen cycling, but as a byproduct of anaerobic fermentation, methane is produced (Moss et al., 2000). Although methane production is a concern for global warming, ruminant animals are essential for the production of nutrient-dense foods from low-quality feedstuffs. This article will discuss rumen microbial ecology, its importance for ecosystem functioning, its contribution to climate change, and its importance for sustainable agriculture.

DOI: https://doi.org/10.5281/zenodo.19607530

An Intelligent Poultry Farm Management System Using Iot And Cloud Based Data Analytics

Authors: S.Senthazhai, V.Kokila, R.Dharshini, B.Pragathi, E.Sonashriyaa

Abstract: This paper presents the design and implementation of a smart environmental monitoring and control system using the Raspberry Pi Pico W microcontroller. The proposed architecture integrates multiple sensors—including temperature and humidity, gas level, water level, and feeder level—to continuously monitor ambient conditions. A forecasting module enhances system intelligence by predicting short-term environmental trends based on real-time data. The Raspberry Pi Pico W processes sensor inputs and communicates wirelessly with a cloud database, enabling remote access via mobile or desktop interfaces. Relay-controlled actuators such as a heater, cooling fan, exhaust fan, water pump, and servo motor respond dynamically to sensor thresholds, ensuring automated regulation of the environment. The system demonstrates a scalable and cost-effective solution for applications in smart agriculture, pet care, and automated home ecosystems. Experimental results validate the system’s responsiveness and reliability, highlighting its potential for real-world deployment in IoT-based automation frameworks.

DOI: https://doi.org/10.5281/zenodo.19608972

Algorithmic Management In Greenhouse Operations: Opportunities, Risks, And Ethical Challenges

Authors: MD Jaynul Abedin, Md Tayef Shiham

Abstract: Controlled-environment agriculture is rapidly becoming data-intensive and cyber-physical with the rapid digitalization of controlled-environment greenhouses. With artificial intelligence, IoT frameworks, and robotic surveillance systems becoming integrated in greenhouse operations, algorithms are playing a larger role in the managerial decision-making process instead of human supervisors alone. This change opens the idea of algorithmic management to the world of agricultural workforce – a field that has not been sufficiently investigated in the existing studies. This paper constructs a socio-technical system to examine the effect of the algorithm systems on workforce scheduling, performance tracking, and coordination of operations in the greenhouse environment. An optimization model in mathematics is presented to structure task distribution based on efficiency, fairness, and worker fatigue where multi-objective scheduling can be used to achieve productivity and human well-being. The paper offers a proposed structured simulation dataset and a survey instrument to help assess worker perceptions of surveillance and autonomy and fairness to support future empirical research. A comparative analysis of traditional and algorithmic management models indicates that there are trade-offs between agency and precision of operations and labor. The results emphasize that algorithmic management in the agricultural sector is not an issue of technical improvement but a governance problem that needs to be transparent, accountable, and human-centered. This study forms a conceptual and analytical base of ethically responsible AI-based workforce management in smart greenhouse settings and adds to the discussion on the future of human-AI collaboration in industrial systems.

DOI: https://doi.org/10.5281/zenodo.19614456

Aesthetiq: AI-Powered Aesthetic Analysis And Personalized Styling Recommendation System

Authors: Sree Vishal G, Sarika K, Dr. K. Geetha

Abstract: The rapid advancement of digital technologies and the widespread use of social media platforms have significantly influenced the way individuals present themselves in modern society. Personal appearance, grooming, and aesthetic presentation have become essential aspects of self-expression and identity. However, selecting appropriate styles, outfits, and visual themes that align with individual preferences and current trends remains a complex and time-consuming task. Traditional approaches rely heavily on manual browsing, personal judgment, and external opinions, which often lack accuracy, consistency, and personalization.To address these challenges, this paper presents Aesthetiq, an Artificial Intelligence-based aesthetic analysis and personalized styling recommendation system. The proposed system is designed as a web-based application that leverages machine learning techniques to analyze user inputs such as facial images, style preferences, and visual attributes. The system performs preprocessing, feature extraction, and classification to identify suitable aesthetic categories and generate personalized recommendations.The architecture of the system consists of a frontend interface for user interaction, a backend server for processing and communication, and an AI module for performing analysis. The database stores user data, input images, and analysis results to enable efficient retrieval and history tracking. The system ensures real-time processing and provides visually interpretable outputs through an interactive dashboard.Experimental evaluation indicates that the proposed system achieves improved accuracy and performance compared to traditional methods. The system enhances user decision-making, reduces effort, and provides tailored recommendations that align with individual preferences and modern trends.

DOI: https://doi.org/10.5281/zenodo.19614962

InstaCraft: A Lightweight CMS For Instagram‑Based Handicraft Businesses

Authors: Rahul Suthar, Navneet Kumar Singh

Abstract: Instagram is widely used by handicraft sellers be- cause it offers a visual showcase for their products. However, Instagram lacks a proper product catalogue and easy inventory tracking. This paper proposes InstaCraft, a simple web CMS built with the MERN stack (MongoDB, Express.js, React.js, Node.js). InstaCraft provides a responsive product catalogue, a basic admin interface for CRUD operations on products, and direct messaging links (WhatsApp click-to-chat and Instagram profile links). The frontend is designed for mobile devices, and images are optimised (resized, converted to WebP, lazy-loaded) to reduce load times. The paper includes TikZ diagrams of the system architecture and data flow, along with example REST API and schema code. Preliminary testing indicates that InstaCraft pages load approximately 30% faster than a raw Instagram feed displaying similar content. Feature tables compare In- staCraft, Instagram-only selling, and full e-commerce platforms. InstaCraft offers a practical middle ground: it requires less effort than a full online store while providing more organisation than Instagram alone.

Sentiment Analysis Of Indian Amazon Product Reviews: A Comparative Study Of Machine Learning And Lexicon-Based Approaches

Authors: Sachin Kalmani, Pratham Shinde

Abstract: Sentiment analysis has become an essential technique for extracting actionable insights from user-generated content on e-commerce platforms. This study presents a comparative analysis of machine learning and lexicon-based approaches for sentiment classification of Amazon India product reviews, where sentiment labels are derived from user star ratings. Three machine learning models — Naive Bayes, Logistic Regression, and Random Forest — are evaluated alongside two lexicon-based methods, VADER and TextBlob. Text preprocessing and feature extraction are performed using standard natural language processing (NLP) techniques combined with TF-IDF vectorization. Models are tested under four train-test split configurations (80-20, 60-40, 40-60, and 20-80) to systematically assess the effect of training data size on performance. Results show that machine learning models consistently outperform lexicon-based approaches across all evaluation metrics. At the 80-20 split, Random Forest achieves the highest accuracy of 96.22%, followed by Logistic Regression at 85.97% and Naive Bayes at 80.58%. Lexicon-based methods plateau near 73-74% accuracy across all split configurations, confirming their insensitivity to training data volume. A notable finding is that at reduced training sizes, Naive Bayes (69.65% at 20-80) underperforms both VADER (74.11%) and TextBlob (72.90%), suggesting that lexicon-based methods are more reliable when labelled training data is scarce. These findings offer practical guidance for model selection in real-world sentiment analysis applications.

DOI: https://doi.org/10.5281/zenodo.19625686

A Novel Ensemble Machine Learning Method To Detect Phishing Attack

Authors: Dr. Pawan Bhaladhare, Vaibhav Ingle, Sakshi Phatake, Rushi Jagtap

Abstract: The rapid growth of internet users has led to an increase in phishing attacks, where attackers create deceptive URLs to steal sensitive information. This study presents an ensemble machine learning framework for detecting phishing websites using Natural Language Processing (NLP) and multiple classifiers, including Artificial Neural Networks (ANN), Naïve Bayes (NB), Random Forest (RF), and Support Vector Machines (SVM). By extracting key features from URLs and applying machine learning techniques, the proposed model enhances detection accuracy. Comparative analysis demonstrates its effectiveness, achieving 98.4% accuracy in distinguishing phishing sites from legitimate ones. This approach offers a proactive solution to mitigate online security threats and protect users from cyber fraud. Phishing attacks have become more sophisticated, using deceptive URLs to target unsuspecting users. This research introduces a hybrid machine learning-based detection model that enhances accuracy through an ensemble of classifiers. The system utilizes Natural Language Processing (NLP) to extract critical URL features, which are then analyzed using Artificial Neural Networks (ANN), Naïve Bayes (NB), Random Forest (RF), and Support Vector Machines (SVM). Machine learning techniques are particularly effective in detecting zero-hour phishing attacks and adapting to emerging threats. Our implementation achieved a 98.4% accuracy in classifying websites as phishing or legitimate.

Brain Tumor Detection and Classification Using Machine Learning

Authors: Dr. A.P Srivastava, Sanjivani sharma, Mayank Kumar Singh, Aman, Saurabh Yadav, Akhand Pratap Vishwakarma

Abstract: Brain tumors are among the most critical neurological disorders and require early and accurate diagnosis to improve patient survival rates. Traditional methods of tumor detection rely heavily on manual analysis of medical images such as Magnetic Resonance Imaging (MRI), which can be time-consuming and prone to human error. This study presents a machine learning–based approach for the automated detection and classification of brain tumors from MRI images. The proposed system utilizes image preprocessing techniques to enhance image quality and remove noise, followed by feature extraction to identify significant patterns associated with tumor regions. Various machine learning algorithms, such as Support Vector Machines (SVM), Random Forest, and Convolutional Neural Networks (CNN), are applied to classify MRI images into tumor and non-tumor categories, and further categorize tumor types. The model is trained and evaluated on a labeled MRI dataset to ensure accuracy and reliability. Experimental results demonstrate that the proposed method improves diagnostic efficiency and achieves high classification accuracy compared to traditional approaches. This automated system can assist radiologists and healthcare professionals in early tumor detection, reducing diagnosis time and improving treatment planning.

DOI: https://doi.org/10.5281/zenodo.19626168

Centralized Automated Solution for Price Estimation and Reasonability

Authors: Thakur Bhargavi Walmik, Khatate Manasvi Santosh, Prof. K. R. Metha

Abstract: The rapid growth of internet users has led to an increase in phishing attacks, where attackers create deceptive URLs to steal sensitive information. This study presents an ensemble machine learning framework for detecting phishing websites using Natural Language Processing (NLP) and multiple classifiers, including Artificial Neural Networks (ANN), Naïve Bayes (NB), Random Forest (RF), and Support Vector Machines (SVM). By extracting key features from URLs and applying machine learning techniques, the proposed model enhances detection accuracy. Comparative analysis demonstrates its effectiveness, achieving 98.4% accuracy in distinguishing phishing sites from legitimate ones. This approach offers a proactive solution to mitigate online security threats and protect users from cyber fraud. Phishing attacks have become more sophisticated, using deceptive URLs to target unsuspecting users. This research introduces a hybrid machine learning-based detection model that enhances accuracy through an ensemble of classifiers. The system utilizes Natural Language Processing (NLP) to extract critical URL features, which are then analyzed using Artificial Neural Networks (ANN), Naïve Bayes (NB), Random Forest (RF), and Support Vector Machines (SVM). Machine learning techniques are particularly effective in detecting zero-hour phishing attacks and adapting to emerging threats. Our implementation achieved a 98.4% accuracy in classifying websites as phishing or legitimate.

DOI: https://doi.org/10.5281/zenodo.19626172

Secure Voting System Using Blockchain Technology: A Decentralized Approach to Enhance Electoral Integrity

Authors: Suraj Yadav, Sagar Gupta, Tanishq Raj Mahaur, Mr. Anurag Anand Duvey

Abstract: The conventional electronic voting machines often have problems like, centralized vulnerabilities, lack of transparency, prone to single-point-of-failure attacks, as well as high administrative overhead for voter verification. This paper presents a solution: A Decentralized e-voting framework created on the Ethereum Virtual Machine (EVM) that tackles these issues through the integration of blockchain immutability and mobile-native biometric authentication. With the medium this project, we propose a system that implements Solidity smart contracts to manage election lifecycles and a React Native frontend for User friendly UI and cross-platform accessibility. The key innovations such as a cycle-based state management mechanism for optimizing the contract reusability and a cryptographic credential-hashing protocol that safeguards voter identity without the need for high-cost third-party verification services.

AI Powered Smart Urban Infrastructure

Authors: Anushka Rastogi, Priya Gupta

Abstract: Urban areas are rapidly expanding and this brings with a set of complicated problems. Things like traffic jams, rising energy bills, overflowing landfills, and public safety concerns are becoming everyday issues for city residents. This paper looks at how artificial intelligence can play a practical role in fixing these problems. AI is making everything rapid and faster, from managing road signals to making everything work wisely and securely. At the same time, the paper does not ignore the hurdles, like data security, the cost of setting up these systems, and making sure that benefits reach every part of society, not just wealthy neighbourhoods. This study also examines how artificial intelligence can help cities better prepare for long-term challenges like climate change, population growth, and natural disasters. By looking at existing AI projects around the world, this research aims to give a realistic view of the current state and potential of AI in city infrastructure.

DOI: https://doi.org/10.5281/zenodo.19630016

Machine Learning And Deep Learning Techniques For Automated Skin Cancer Detection: A Comprehensive Review

Authors: Shruti Chouhan, Prof. Pankaj Raghuwanshi

Abstract: Skin cancer is one of the most prevalent and rapidly increasing forms of cancer worldwide, making early detection essential for improving patient survival and treatment outcomes. Traditional diagnostic methods rely heavily on visual examination and dermoscopic analysis by dermatologists, which may sometimes be subjective and dependent on clinical expertise. In recent years, machine learning (ML) and deep learning (DL) techniques have emerged as powerful tools for automated skin cancer detection and classification. These techniques utilize medical image datasets, particularly dermoscopic images, to identify patterns and features associated with malignant and benign skin lesions. This review presents a comprehensive analysis of recent research on ML and DL-based approaches for automated skin cancer detection. Various algorithms such as Support Vector Machines (SVM), Random Forest, Convolutional Neural Networks (CNN), and transfer learning models are examined in terms of their methodologies, datasets, and performance metrics. Additionally, this study highlights the advantages, limitations, and challenges associated with these techniques. The review also discusses future research directions, including the development of more diverse datasets, interpretable models, and integration of AI-based systems into clinical practice to enhance diagnostic accuracy and healthcare efficiency.

From Words To Intelligence: A Comprehensive Survey Of Large Language Models And Their Transformative Role In Natural Language Processing

Authors: Sai Rithwik Nooguri

Abstract: The emergence of Large Language Models (LLMs) represents one of the most consequential shifts in the history of artificial intelligence (AI) and natural language processing (NLP). Built on the Transformer architecture with self-attention mechanisms, LLMs such as BERT, GPT-3, T5, LLaMA, and GPT-4 have achieved state-of-the-art performance across a broad spectrum of linguistic tasks, fundamentally reshaping how machines comprehend and generate human language. This survey presents a systematic and comprehensive review of the evolution of NLP—from rule-based and statistical methods to the current era of foundation models—examining key architectural innovations, pre-training objectives, fine-tuning strategies including parameter-efficient methods such as Low-Rank Adaptation (LoRA), and alignment techniques including Reinforcement Learning from Human Feedback (RLHF). We critically assess performance across standard benchmarks including GLUE, SuperGLUE, and MMLU, and analyze persistent challenges such as hallucination, bias, computational cost, and explainability. Furthermore, we explore the expanding landscape of LLM applications in healthcare, education, legal reasoning, and code generation, and outline promising future directions including multimodal models, efficient inference, and AI alignment. This work aims to serve as both an accessible introduction and a scholarly reference for researchers and practitioners engaged with the rapidly evolving frontier of AI-powered language understanding.

DOI: https://doi.org/10.5281/zenodo.19630886

Comparative Analysis Of Basic Supervised Machine Learning Algorithms For Iris Flower Classification

Authors: Abu Aasim

Abstract: The Iris Flower Classification problem is one of the most fundamental and widely studied benchmarks in supervised machine learning. It involves classifying iris flowers into three species (Setosa, Versicolor, and Virginica) based on four morphological features: sepal length, sepal width, petal length, and petal width. This review paper clearly defines the category of basic supervised machine learning tasks and explores the existing algorithms for classification. A novel comparative framework is proposed using Python and scikit-learn to evaluate five basic supervised algorithms—Logistic Regression, K-Nearest Neighbors (KNN), Support Vector Machine (SVM), Decision Tree, and Naive Bayes—on the UCI Iris dataset. Performance is measured using accuracy, precision, recall, and F1-score. The study demonstrates that while all algorithms achieve high accuracy (>95%), KNN and SVM consistently outperform others in terms of perfect classification on the test set, highlighting their suitability for simple, linearly separable datasets. General Terms: Supervised Machine Learning, Classification, Comparative Analysis, Iris Dataset, Performance Metrics.

Foreign Direct Investment (FDI) In Bangladeshs Automobile Sector: Trends, Challenges, And Policy Implications

Authors: Zannatul Rumman Zinia, Abdullah Al Ruhul

Abstract: This study examines the economic impact of Foreign Direct Investment (FDI) on Bangladesh’s automobile sector, with particular emphasis on sectoral output, employment generation, and macroeconomic determinants of investment inflows. Using annual time-series data and sector-specific indicators, the analysis integrates descriptive statistics, correlation assessment, multiple regression modeling, and iterative epoch-based robustness evaluation to investigate both the contribution and sustainability of FDI-led industrial growth. The empirical results indicate that manufacturing-oriented FDI exerts a positive and statistically significant influence on automobile sector gross value added (GVA), supporting the hypothesis that foreign capital contributes to capital deepening, technology diffusion, and production expansion. Real GDP, serving as a proxy for market size, emerges as a strong determinant of FDI inflows, while human capital development and trade openness demonstrate complementary roles in enhancing investment attractiveness. However, the employment elasticity of FDI remains moderate, suggesting that capital-intensive investment patterns dominate labor absorption effects. Productivity growth, measured as output per worker, exhibits gradual improvement but reflects structural constraints related to limited local value-chain integration. The findings suggest that while FDI plays a constructive role in supporting sectoral expansion, its long-term developmental impact depends on institutional quality, skill upgrading, and domestic supplier ecosystem strengthening. Policy recommendations emphasize targeted human capital development, enhanced local content integration, regulatory efficiency, and export-oriented industrial clustering to maximize the transformative potential of manufacturing FDI within Bangladesh’s automobile industry.

DOI: https://doi.org/10.5281/zenodo.19640120

PDF Summarization And Query Answering: A Hybrid AI-Driven Approach

Authors: D.Hari Priya, Ch.Charmi Sri, A.Rohit, K.Harika Sri, Ms. M. Soumya

Abstract: This paper presents PDFChatBot, a comprehensive AI-driven system for automated PDF summarization and intelligent query answering. Our hybrid approach integrates Rhetorical Structure Theory (RST), transformerbased models (BERT, GPT-4, Gemini-1.5-Pro), and FAISS vector databases, achieving state-of-the-art ROUGE-L scores of 0.51 and F1-scores of 0.87 across 50 diverse documents spanning research papers, legal contracts, medical reports, financial statements, and technical manuals. The system processes 100-page documents in under 120 seconds, reducing document review time by 80% while maintaining semantic coherence. We demonstrate superior performance over TextRank (ROUGE-L: 0.37), BART-large (0.44), and T53B (0.47) baselines through rigorous evaluation across five distinct domains. Production-ready deployment via FastAPI, Streamlit, Docker, and Redis caching ensures scalability for enterprise applications with 99.9% uptime and sub-second query latency.

DOI: https://doi.org/10.5281/zenodo.19642249

Student Performance Indicator: An End-to-End Machine Learning Pipeline for Predicting Academic Outcomes

Authors: Smit Sudani

Abstract: With all the amount of data that is now available about the students in a school environment, there is no way one could analyze such data manually. The Student Performance Predictor is a web application I designed to help determine the final score that a particular student will get from mathematics class, basing on his demographics and background. The whole machine learning pipeline was implemented by me using the Python language. After experimenting with various models in Jupyter Notebooks and having my kernel crash quite a few times, I managed to find the most accurate one – Random Forest Regressor with an 80% accuracy rate. Next, I embedded this algorithm in my application, which uses the Flask server. User only needs to input some values in three fields to get the prediction instantly.

Learning Management System Using Web Technology

Authors: Ansari Zain, Khan Fahad, Rajput Burhan, Khan Shifa, Chandramohan Konduri

Abstract: A Learning Management System (LMS) is a comprehensive web-based application developed to streamline the process of teaching, learning, and academic administration. The main objective of the LMS is to provide a unified digital platform where educators can create, organize, and manage learning content, while learners can easily access courses, participate in discussions, submit assignments, and track their academic progress. The system eliminates geographical and time limitations, enabling flexible and self-paced learning for students across different devices. The proposed LMS includes essential modules such as user authentication, course management, content uploading, online assessments, grading, progress tracking, and communication tools like notifications and discussion forums. It leverages database management systems to securely store and retrieve user data, ensuring reliability and scalability.

DOI: https://doi.org/10.5281/zenodo.19642420

The Impact Of Artificial Intelligence On Cybersecurity

Authors: Rathod Alfaz, Ravi Ranjan Kumar Pandey

Abstract: Artificial Intelligence (AI) has changed many industries, and its influence on cybersecurity is very significant. This research paper studies the progress of AI and its role in handling the changing challenges of cybersecurity. It examines the possible benefits of AI in threat detection, vulnerability assessment, incident response, and predictive analytics. In addition, the paper discusses the ethical concerns and possible risks connected with AI in cybersecurity. Through the study of current research, case studies, and industry practices, this paper aims to provide clear insights into the opportunities and challenges created by the use of AI in the field of cybersecurity.

Medico: Design, Development, And Validation of a Scalable Web-Based Platform for Digital Healthcare Appointment Management

Authors: Prof. Biju Balakrishnan, Deep Patel, Pankitkumar Patel, Virajkumar Suthar, Dharmik Kanojia

Abstract: Healthcare delivery in its conventional form continues to face persistent operational hurdles — prolonged patient waiting periods, excessive administrative burden, and geographic constraints. This paper introduces MEDICO, a security-focused and patient-centred web portal engineered to establish a fluid digital healthcare environment. The platform was built using the Python Django framework, adopting a Waterfall development methodology and implementing a Model-View-Template (MVT) architecture to support dual-role access control and an intelligent appointment scheduling engine. System capabilities — including practitioner discovery, profile browsing, and automated notification dispatch — were evaluated through Unit, Integration, and System-level testing. Quantitative stress testing demonstrated complete transactional integrity while concurrently processing 50 simultaneous appointment requests, recording zero system failures or scheduling conflicts. This lays a dependable technical foundation directly combating the inefficiencies of conventional booking methods. Subsequent development phases will focus on Artificial Intelligence (AI) for personalised doctor recommendations and a fully integrated Electronic Health Record (EHR) management module.

DOI: https://doi.org/10.5281/zenodo.19642551

Storytales : Ai Tells The Story Automated Story-To-Video Generation Using Generative Artificial Intelligence

Authors: Manasi Rathod, Jayesh Mahajan, Rutwij Landge

Abstract: Storytelling represents one of the most effective techniques for communication, education, and knowledge transfer across diverse domains. However, traditional text-based storytelling methods often fail to maintain engagement among modern learners who increasingly prefer visually rich multimedia experiences. Creating animated storytelling videos manually requires expertise in scripting, illustration, animation design, narration recording, and editing tools. This paper presents STORYTALES – AI Tells the Story, an automated Generative Artificial Intelligence framework that converts textual narratives into animated storytelling videos with synchronized narration and scene-wise visualization. The system integrates Large Language Models for semantic scene segmentation, Stable Diffusion XL for visual synthesis, Stable Video Diffusion for animation generation, Coqui XTTS for narration synthesis, and FFmpeg for automated multimedia composition. Experimental evaluation confirms that the proposed architecture significantly reduces multimedia production complexity while improving accessibility for educators and content creators.

DOI: https://doi.org/10.5281/zenodo.19644537

Customization of Time Slots for Delivery of Articles and parcels using Artificial Intelligence

Authors: Soumya M Achari, Pakhi Singha, Nithin Ramakrishnan

Abstract: On demand delivery began as a competitive edge in the consumer market. Quick commerce sites provided customers access to products within the shortest possible time to stand out from rivaling brands. However, this fast growth of 10 minute and 1 day delivery services leave traditional delivery services irrelevant. Due to the customer’s opting for convience and speed, retailers selling stock struggle to meet these expectations and lose profitability. Access to real time data updates and optimisation has hence become significant in ensuring delivery to correct locations, punctually and efficiently. Current local systems struggle to respond to dynamic data, leading to missed delivery time slots, manual intervention requirement, excessive fuel and time wastes, poor customer feedback and so on. In order to remain competitive in such consumer markets, business require real time updates on demand and supply chains, delivery agent availability, client shopping patterns and traffic volume information. To counter these challenges artificial intelligence can be used to understand real time data and set parcel delivery time slots automatically while routing delivery agents through optimal pathways and monitoring the system of the agents and customer to align with their available schedules. The AI will utilise previous ETA, traffic congestion, pattern recognition in relation to prior on time articles that were received and user presence to define schedules for delivery and update the consumers, drivers and supervisors accordingly. This proposed intelligent system would solve the common E-Commerce problems faced by traditional delivery systems by ensuring routes are mapped to avoid redundancy, increase time efficiency, deliver as per consumer availability, especially for cash on delivery where the client is required at the home for payment, provide real time transportation status of the products to supervisors and customers, therefore increasing the trust of the user in the brand and providing an avenue for the manager to handle mismanaged deliveries. Such a system would bolster customer satisfaction and also reduce fuel and time consumption for the drivers, enhancing their work life balance. Deliveries that are more likely to be missed or routes that could result in accidents would be information sent to the supervisor, customer and delivery agents respectively, hence, preventing missed deliveries, injuries and delays. Such systems have been applied experimentally at a smaller scale and proven successful in reducing time, fuel, costs and injury risk, while improving customer satisfaction, making them a worthwhile subject of research.

DOI:

Deep Learning Based Classification of Liver Diseases Using Heterogeneous Ultrasound Image

Authors: Anto Maurin Lisha L, Muthu M, Sadeesh P, Tamilarasan S

Abstract: Liver diseases such as fatty liver, cysts, and tumors require early and accurate diagnosis to improve patient outcomes. Ultrasound imaging is widely used due to its non-invasive and cost-effective nature; however, its heterogeneous characteristics, including speckle noise, low contrast, and variability across devices, make diagnosis challenging. This paper proposes a deep learning-based approach for the classification of liver diseases using heterogeneous ultrasound images. The system employs pre-processing techniques such as noise reduction, normalization, and contrast enhancement to improve image quality. A YOLO-based architecture integrated with convolutional neural networks is used for feature extraction and simultaneous detection and classification of liver abnormalities. Experimental results show that the proposed model achieves improved accuracy and robustness compared to conventional methods. The system supports real-time analysis and can assist clinicians in reliable and efficient liver disease diagnosis.

Arduino-based Firefighting Robot

Authors: Dr. Ch. Venkata Krishna Reddy, B. Varun Tej, T. Prabhas, G. Vishnu Charan

Abstract: Accidents caused by fire result in severe damage to life and property, especially in hazardous and hard-to-reach areas. In order to minimize human risk and increase the efficiency of firefighting, a Fire Fighting Robot with ESP32 Camera is proposed and implemented. In this system, the Arduino Uno board is used as a primary controller. The ESP32-CAM is used for live video streaming through a web page for the user. The robot is designed to operate in two modes: manual mode and automatic mode. The modes are selected through a web page. In manual mode, the user controls the robot’s movement and views the live video feed. The ultrasonic sensor is used in manual mode for obstacle detection. Four flame sensors are used to detect fire. Once the fire is detected, the robot moves towards the fire source. A DC water pump is used to spray water on the fire and extinguish it. The robot’s movement is controlled using DC motors driven by an L298 motor driver. A servo motor is used for direction control of the water pump. A buzzer is used for alarm generation when the fire is detected. The robot is powered using a battery supply regulated using an LM2596 voltage regulator module. This project is a simple and cost-effective way of remote fire detection and firefighting using robotics and wireless monitoring techniques. It is useful for industrial areas, warehouses, and places where human access is hazardous.

DOI: https://doi.org/10.5281/zenodo.19658551

Impact Of Artificial Intelligence On Consumer Behavior

Authors: Vansh Nigam, Mr. Pankaj Lalwani

Abstract: Artificial Intelligence (AI) is no longer just a futuristic concept; it has quietly become a part of our daily lives, influencing the way people search, shop, and interact with brands. From personalized recommendations on e-commerce platforms to virtual assistants answering queries in real time, AI has started to reshape how consumers make decisions. This research paper focuses on understanding the impact of AI on consumer behaviour, looking beyond the technology itself to explore how it changes trust, buying patterns, loyalty, and expectations. The study examines how AI creates value by offering convenience and personalization—consumers now expect brands to “know them” and provide solutions tailored to their needs. At the same time, it highlights challenges such as privacy concerns, over-reliance on algorithms, and the risk of losing the human touch in brand–consumer relationships. By analysing existing studies, market practices, and consumer perceptions, this paper aims to bridge the gap between technological advancement and human psychology. Ultimately, the research argues that AI is not just influencing consumer choices but also shaping a new kind of consumer—more informed, more connected, and more demanding. Businesses that can balance AI-driven efficiency with ethical responsibility and genuine human engagement will be the ones to build lasting trust in the age of intelligent technology.

DOI: https://doi.org/10.5281/zenodo.19658824

Hospital-Based Smart Hematology Analyzer with Cancer Risk Alert

Authors: Aarthi R, Ranjith S, Subash P, Surya T

Abstract: The Hospital-Based Smart Hematology Analyzer with Cancer Risk Alert is an advanced system designed to automate blood analysis while providing early cancer risk detection for organs such as the brain, lung, and skin. The system integrates a deep learning algorithm, InceptionV3, to analyse blood smear images and identify abnormal cell patterns indicative of potential malignancies. High-resolution images captured through an optical sensor are pre-processed and fed into the algorithm for feature extraction and classification. The hardware architecture includes a microcontroller interfaced with sensors and a display unit, interconnected through UDP communication to ensure fast, reliable, and real-time data transfer within the hospital network. The analyser automatically computes hematology parameters such as RBC, WBC, haemoglobin levels, and platelet count, while the AI module evaluates potential cancer risk based on morphological anomalies. Alerts and reports are generated for medical staff if any abnormal patterns are detected, facilitating prompt medical intervention. The working flow begins with blood sample collection, followed by automated slide preparation, image acquisition, and pre-processing. The processed images are analysed by the InceptionV3 model, which classifies the results and calculates risk levels. Data is transmitted via UDP to a central monitoring system for visualization, record keeping, and further evaluation by doctors. This system emphasizes automation, real-time analysis, and predictive diagnostics, aiming to reduce manual errors, accelerate clinical decision-making, and improve early cancer detection. It provides a cost-effective, intelligent, and scalable solution for hospital-based patient care.

Smart Playlist Generator Using Affective Computing

Authors: Drbrindhas, Ms. P.Abirami In, Mr. Ajay.R, Mr. Anbarasan.R, Mr. Rishihesh .M.M, Mr.Safwan.S, Mr.Sriram.V

Abstract: This paper presents the design and implementation of a Smart Playlist Generator using Affective Computing — a real-time, AI-driven music recommendation system that personalizes playlists based on the user’s emotional state. The system integrates three core components: (1) a Facial Emotion Recognition (FER) module built on OpenCV and Convolutional Neural Networks (CNNs) that classifies emotions in real time from webcam input, (2) a Natural Language Processing (NLP) module that supports Thanglish (Tamil- English transliterated) text commands for conversational interaction, and (3) a Spotify Web API integration that maps detected emotions to audio features such as valence, energy, and tempo to generate context-aware playlists. The system achieves an emotion recognition accuracy of 87– 90%, Thanglish command interpretation accuracy exceeding 90%, and a playlist-mood alignment rate of 85–90%, with an end-to-end latency of approximately 3 seconds. The architecture leverages HTML/CSS/JavaScript for the frontend, Node.js with Express for the backend, Firebase for data persistence, and Python-based AI modules for emotion and language processing. Experimental results confirm the viability of affective computing for dynamic, personalized music delivery, and the system demonstrates significant potential for next- generation human-computer interaction in multimedia platforms.

DOI: https://doi.org/10.5281/zenodo.19659822

AI-Integrated Android And Mobile Development Framework Mahesh Saini & Guided By Dinesh Cholkar

Authors: Mahesh Saini, Dinesh Cholkar

Abstract: The rapid evolution of mobile computing has fundamentally transformed how humans interact with technology. This paper presents an AI-Integrated Android and Mobile Development Framework (AI-AMDF) that leverages machine learning, cross-platform development tools, and intelligent UI/UX systems to deliver high-performance, adaptive mobile applications. The proposed framework dynamically optimizes app behavior, battery usage, and user experience based on real-time device analytics and user interaction patterns. Results demonstrate a 38% improvement in app performance metrics and a 31% reduction in development time-to-deployment compared to conventional mobile development approaches.

AuctionOasis: A Scalable Web-Based Platform For Real-Time Live Auctions

Authors: Yash Sakhareliya

Abstract: Auction systems have become increasingly popular as the uptake of e-commerce grows globally. However, conventional auction systems may be inflexible and unable to accommodate several bidders simultaneously. To address these issues, AuctionOasis provides a modular and comprehensive web platform that incorporates real-time bid processing, auction management, and secure participation of users. The platform is developed using Node.js, Express.js, MongoDB, and EJS for front-end rendering.Further, the system is planned to implement the Socket.io technology for conducting group live bidding and chatting. This document discusses the motivation behind developing AuctionOasis, its architectural framework, design aspects, implementation process, and future directions.

AI-based Cyber Threat Prediction Framework

Authors: Mohit Japee, Parthi Soni

Abstract: Modern enterprise networks generate a large volume of security events, making it difficult for security analysts to identify critical threats in real time. Traditional rule-based detection mechanisms often fail to detect advanced and evolving cyber attacks. Artificial Intelligence (AI) and Machine Learning (ML) techniques have shown promising capabilities in analyzing large-scale security data and predicting potential cyber threats. This research proposes an AI-based cyber threat prediction framework designed to enhance threat detection and decision-making in enterprise environments. The framework focuses on log analysis, anomaly detection, and threat prediction using machine learning techniques. The study highlights the potential of predictive analytics in improving proactive cybersecurity strategies and reducing response time in security operations centers (SOCs). The proposed framework is conceptual and aims to provide a cost-effective and scalable approach for organizations adopting intelligent cybersecurity solutions.

DOI: https://doi.org/10.5281/zenodo.19675247

Shop Gara: A Complete E-Commerce Solution

Authors: Mohammad Atiullah Ansari

Abstract: Shop Gara is a modern digital platform designed to facilitate seamless cross-border trade for Nepalese businesses and consumers. The platform streamlines international procurement and sales by offering secure transactions, efficient logistics, and transparent trade procedures. This paper presents the design, development, testing, and future scope of Shop Gara — a scalable, reliable, and efficient cross-border e-commerce platform. It also highlights the impact of digiti- zation on Nepal’s international trade landscape.

DOI: https://doi.org/10.5281/zenodo.19661712

Design And Development Of Multi-Source Renewable Energy Integrated EV Charging Station

Authors: Dr Ch V Krishna Reddy, MVV Aneesh, V Yushma Naga Surya, N Akhil Sai

Abstract: The dependence of electric vehicle (EV) charging on grid electricity drawn from fossil-fuel-based generation defeats the purpose of sustainable mobility. This paper presents the design and prototype development of a tri-hybrid renewable EV charging station that simultaneously harvests energy from three independent sources: a Solar Photovoltaic (PV) array, a Vertical Axis Wind Turbine (VAWT), and a Regenerative Speed Bump (RSB) mechanism. Each source feeds a common 12 V DC bus through dedicated signal-conditioning chains — MPPT for solar, and full-bridge rectifier followed by a voltage regulator for VAWT and RSB — with isolation diodes preventing reverse current flow. The conditioned power charges a 12 V, 5 Ah lead- acid battery, from which an EV charging dock is supplied. Energy calculations based on actual prototype hardware yield a gross daily harvest of approximately 182.9 Wh, reducing to roughly 128 Wh of usable energy after system losses. The three sources are chosen for their complementary availability: solar dominates during clear daylight, wind supplements during low- irradiance periods, and the speed bump produces a transient energy pulse with every vehicle passage. The prototype, demonstrates a cost-effective, grid-independent charging solution scalable to full-station deployment.

DOI: https://doi.org/10.5281/zenodo.19678780

Investor Awareness And Preference Toward Mutual Funds In India

Authors: Nitish kumar, Dr Sajad Ahmad Mir

Abstract: Mutual funds have emerged as one of the most relevant investment avenues for Indian retail investors because they provide diversification, professional management, flexibility, and accessibility through Systematic Investment Plans (SIPs). This study examines investor awareness and preference toward mutual funds in India, with special reference to retail investors. It focuses on the level of awareness, preferred investment avenues, mutual fund category preference, SIP adoption, and the key barriers that prevent wider participation. The study is based on a descriptive and analytical survey design using primary data collected from 100 respondents through a structured questionnaire. Secondary sources from regulatory bodies, industry reports, and academic literature were also used to strengthen the discussion. The findings show that awareness of mutual funds is moderate to high among the respondents, SIPs are the dominant investment mode, and equity mutual funds are the most preferred category. Mutual funds were also identified as the most preferred investment avenue in the sample, ahead of fixed deposits and gold. At the same time, lack of knowledge, fear of market risk, and limited surplus income remain major barriers to investment. The study concludes that awareness alone does not guarantee participation; investor confidence, financial literacy, and simple communication are equally important. Practical recommendations are offered for investors, mutual fund companies, and regulators to improve informed participation in the Indian mutual fund market.

Transforming The Human In Human Resources: Artificial Intelligence As A Catalyst In Strategic HR Decision-Making

Authors: Prakash Kumar Sinha, Dr. Navneet Seth

Abstract: Artificial Intelligence (AI) is rapidly transforming the field of Human Resource Management (HRM) by redefining traditional practices and enabling data-driven strategic decision-making. This study explores the role of AI as a catalyst in strategic HR decision-making and examines how intelligent technologies are reshaping the “human” aspect of human resources. The research highlights the integration of AI-powered tools such as predictive analytics, machine learning, chatbots, and automated recruitment systems in key HR functions including talent acquisition, employee engagement, performance evaluation, workforce planning, and retention management. The study emphasizes that AI enhances organizational efficiency by reducing manual workload, minimizing bias in recruitment processes, improving accuracy in decision-making, and enabling personalized employee experiences. At the same time, it discusses the challenges associated with AI adoption, including ethical concerns, privacy issues, lack of emotional intelligence, and resistance to technological change within organizations. The paper further argues that AI should not replace human judgment but rather complement human capabilities in strategic HR practices. The successful implementation of AI in HR requires a balanced approach that combines technological innovation with empathy, ethical standards, and human-centered leadership. The study concludes that AI has emerged as a powerful strategic partner in HR decision-making, enabling organizations to achieve greater productivity, agility, and competitive advantage in the evolving digital era.

A Study On Diophantine Equations

Authors: Dr Vinit Kumar Sharma, Kanika

Abstract: Diophantine equations are polynomial equations in two or more variables whose solutions are restricted to integers. These equations are named after the ancient Greek mathematician Diophantus of Alexandria, who made significant contributions to number theory through his work Arithmetica. Diophantine equations play an important role in modern mathematics, cryptography, computer science, optimization, and algebraic geometry. Although Diophantine equations are a branch of pure number theory concerned with integer solutions of equations, they have several practical applications that affect everyday human life, often indirectly through technology and optimization.

DOI: http://doi.org/10.5281/zenodo.20503729

Relationship Between Graph Theory and Network Analysis: Foundations, Applications, and Future Directions

Authors: Dr. Vinit Kumar Sharma, Aanchal Gupta

Abstract: Graph theory and network analysis are closely interconnected disciplines that play a crucial role in understanding complex systems. Graph theory provides the mathematical framework for representing entities and relationships, while network analysis applies these concepts to investigate structural and functional properties of real-world networks. This paper explores the relationship between graph theory and network analysis, discusses key graph-theoretic measures, examines applications across multiple domains, and identifies emerging research directions including graph neural networks and temporal networks. The study demonstrates that graph theory serves as the theoretical backbone of network analysis and continues to drive innovations in modern data science and artificial intelligence.

DOI: http://doi.org/10.5281/zenodo.20507775

A Study of Group Theory with Various Properties

Authors: Dr Vinit Kumar Sharma, Insaf Ali

Abstract: Group Theory is one of the fundamental branches of modern algebra and provides the mathematical framework for studying symmetry, transformations, and algebraic structures. This paper presents a comprehensive study of groups and their fundamental properties, including closure, associativity, identity, inverses, commutativity, cyclicity, normality, solvability, and nilpotency. The paper examines finite and infinite groups, discusses important classification theorems, and highlights applications in cryptography, coding theory, physics, and computer science. A graphical analysis is presented to compare major classes of groups according to their structural complexity. Recent developments in finite group classification and graph-theoretic approaches to group theory are also reviewed.

DOI: http://doi.org/10.5281/zenodo.20759238

Feature-Enhanced Deep Learning Framework For Early Detection Of Coconut Leaf Diseases

Authors: R.Kanimozhi, Dr.V.Maniraj

Abstract: Coconut is an economically important plantation crop whose productivity is significantly affected by leaf diseases such as Leaf Rot, Grey Leaf Spot, and Bud Rot. Early detection of these diseases is essential to reduce crop losses and improve yield. Conventional disease diagnosis through manual inspection is time-consuming, subjective, and unsuitable for large-scale plantations. This paper proposes an Attention-Based Deep Learning Framework for the early detection and classification of coconut leaf diseases. The proposed framework integrates image preprocessing, data augmentation, transfer learning, and a Convolutional Block Attention Module (CBAM) to enhance feature extraction. The attention mechanism enables the model to focus on disease-affected regions while suppressing irrelevant background information. A convolutional neural network is used to classify healthy and diseased leaf images with improved accuracy. The model is evaluated using standard performance metrics, including accuracy, precision, recall, F1-score, and confusion matrix. Experimental results demonstrate that the attention-based framework outperforms conventional CNN models in detecting early-stage disease symptoms. The proposed approach is computationally efficient and suitable for real-time deployment on mobile and edge devices. It provides an effective decision-support tool for farmers and agricultural experts, contributing to improved disease management and sustainable coconut cultivation.

DOI: http://doi.org/10.5281/zenodo.21468251