Category Archives: Uncategorized

Authors: Assistant Professor Mrs. B. Aruna Kumari, Immadi Naga Varshitha, Ramadeni Vasavi, Para Prasanthi, Marripudi Jeevana Jyothi

Abstract: One of the essential technologies that allow implementing the human-computer interaction built intuitively and with a certain level of comfort is the recognition of hand gestures, in particular, in the smart home automation systems. This paper presents a new deep learning model, Attention-Enhanced CNN Gesture Recognition (AE-CNN-GR) that can enhance the quality, responsiveness, and resilience of gesture control on live camera streams and improve the accuracy. The model is based on the extension of the traditional CNN architecture, incorporating channel and spatial attention units, to enable the network to concentrate on the most informative parts of the hand, such as fine finger movements and changes of the positions. Channel attention module records finer spectral and intensity differences in parts of the hands and the spatial attention mechanism focuses on important geometric and contextual characteristics of gestures to enhance the accuracy of classification and boundary detection. The methods of transferMediaPipe and OpenCV identifications and preprocessing using hand detection and appliance control with the use of the Arduino simulation.

DOI: https://doi.org/10.5281/zenodo.20649582

Published by: vikaspatanker

Authors: Assistant Professor G. Lakshmi Durga, Gade Bhagya Sri, Guntaka Vahnitha, Shaik Benazil Bhanu, Devarapalli Thanuja

Abstract: The Internet of Things (IoT) technology allows individuals to have new interfaces to communicate with devices that are smart. We present a Smart IoT Interface with Hand Gesture Recognition and Machine Learning in this work to enhance human- machine interaction (HMI) in smart environments. Being a wearable hand gesture recognition and control device, it relies on sensor networks and embedded systems to obtain real-time hand gesture feedback, which is later interpreted by advanced machine learning algorithms to allow natural and natural interaction with IoT devices. The suggested interface takes advantage of wireless communication and edge processing that allows the practical and low-latency processing of real-time data and cloud integration to provide additional device control and gain analytics. Its applications include IoT automation, home automation and an intelligent IoT control to a flexible and reliable system that enables the user to interact with devices connected to it. Findings suggest that the sugg

DOI: https://doi.org/10.5281/zenodo.20649238

Published by: vikaspatanker

Authors: Assistant Professor Srikanth Kilaru, Akireddy Bhargavi, Vadlamudi Bhavitha, Neelam Jyothir Mahitha, Chaparapu Meghana Reddy

Abstract: Smart homes play a crucial role in reducing the amount of energy consumed in the house as the automatic control is also provided. This paper suggested an IoT-based temperature-based fan control system, which is an automatic fan control system that is operated by the temperature in the surrounding. This system contains a LM35 sensor of temperature to detect the temperature with accuracy and the DHT11 sensor to monitor the temperature and humidity in real time. The sensor information is handled in a microcontroller with an ESP8266 Wi-Fi chip that enables the sensor to access the internet easily and visualize the obtained air quality on the cloud. The DC motor speed of the fan is regulated by the Pulse Width Modulation (PWM) which enables your motherboard to provide only the necessary cooling when required. The automatic speed control system makes it non-manually based and the operator is given a pleasant working experience. It has been experimentally established that the proposed variable-speed cooling system

DOI: https://doi.org/10.5281/zenodo.20648914

Published by: vikaspatanker

Authors: Assistant Professor Mrs. G. Rohini Phaneendra Kumari, Ravikrinda Hemanjali,, Manasa Kunduru, Yanamadala Naga Lakshmi,, Chundi Pallavi

Abstract: Increased demand of energy-efficient technologies has resulted in the creation of intelligent systems that would optimize the energy use in residential and commercial buildings. In this paper, the design and development of an automatic light control system of indoor environment that ensures that there is minimal energy wastage through the use of adaptation of illumination is presented. The system makes use of a set of sensors, such as motion sensors and light-dependent resistors (LDRs) to automatically control the lighting through occupancy and the intensity of the ambient light. A framework based on an IoT provides the ability to monitor and control remotely through the use of mobile devices, which makes it more convenient and flexible to the user. The proposed system will provide the optimal lighting conditions and produce a considerable reduction in the electricity consumption, and hence, it will lead to sustainable energy management and smart home automation.

DOI: https://doi.org/10.5281/zenodo.20648720

Published by: vikaspatanker

Authors: SK. Sharmila, Pavuluri Pavani,, Pavuluri Nandini,, Yarramneni Sushma, Onteru Keerthi

Abstract: Safe and potable water must be maintained, and the quality water should be checked frequently, especially due to the pollution and environmental shift. So as to analyse sensor-based method of monitoring water quality, this research integrated temperature reading, total dissolved solids (TDS) and turbidity. The scheme was aimed at the real-time data gathering and evaluation to identify alterations in the water parameters, which will prove the contamination or the quality decline. The results have proven that the combination of input of several sensors enhanced the accuracy and reliability of water quality determination, allowing to identify the possible dangerous situation in time. The paper brings to the fore the possibilities of automated sensor networks in streamlining the water management process and protecting human health

DOI: https://doi.org/10.5281/zenodo.20648454

Published by: vikaspatanker

Authors: Associate Professor P.Sandhya Krishna, Ala Nandini, Pavuluri Sri Lekha, Gumma Aparna, Patchava Pujitha

Abstract: Clean and safe drinking water is a crucial factor in the health of the population, but even now, delivery of contaminated drinking water remains one of the world issues. Water potability: a ML approach The use of ML models in Water Quality Assessment is a recent phenomenon in the past years, as it is now a highly promising tool that predicts the water potability in an efficient (more efficient than traditional) manner. The paper presents a smart machine learning system to anticipate the potability of water that is determined by undertaking a thorough review of diverse physico-chemical characteristics of water such as PH, Hardness, Solids, Chloramines, Sulfate and organic contaminants. State of art preprocessing methods are also applied to address missing values, outliers and feature stratification which enhance the quality and the strength of the data. There are several supervised learning processes, which include Random Forest, SVM, Gradient Boosting and ANN to determine the best predictive accuracy algorithm. The general performance is also justified with the premises of accuracy, precision, recall, F1-score and ROC-AUC performance parameters and demonstrates that the suggested framework implementation is reliable and efficient on actual water quality monitoring scenarios. Also, the work places emphasis on the effects of the feature selection and the hyperparmeter tuning on the enhancement of the prediction performance. Ensemble approach and cross-validation methods cut down on the framework and expand the generalization potential with different datasets.

DOI: https://doi.org/10.5281/zenodo.20648118

Published by: vikaspatanker

Authors: Niravkumar Mahendrabhai Panchal

Abstract: Customer churn has become a major challenge for global e-commerce businesses due to increasing market competition and changing consumer behaviour. This study examines the role of machine learning and big data analytics in predicting customer churn and improving customer retention strategies. A quantitative research design with secondary data sources was adopted to analyse customer behaviour patterns and predictive modelling techniques. The findings indicate that machine learning algorithms and predictive analytics significantly improve churn prediction accuracy and support personalised customer engagement strategies. The study highlights the importance of data-driven decision-making in international e-commerce and provides practical insights for improving customer loyalty, profitability, and long-term business sustainability.

DOI: http://doi.org/10.5281/zenodo.20646827

Published by: vikaspatanker

Authors: Associate Professor V. Pavani, Challagundla Amrutha, Palanati Sirisha, Ganjapu Sowmya, Gottipatti Tejaswini

Abstract: Precise prediction of rainfall is required in agriculture, management of water resources and mitigation of disasters. The nonlinear and uncertain characteristics of the meteorological data are usually difficult to capture by traditional statistical models. As a solution to this, a hybrid stacking ensemble model based on the combination of Random Forest (RF) and Support Vector Machine (SVM) and Logistic Regression as a meta-classifier is proposed. The model, when using the Rain in Australia data set, has the highest accuracy with a value of over 95% in the present version and the possible accuracy of over 96% with superior prepossessing, feature engineering, and class balancing. The suggested method provides a sure model of enhanced rainfall forecasting, which would be involved in planning the sustainability of agriculture and environmental decision-making.

DOI: https://doi.org/10.5281/zenodo.20646431

Published by: vikaspatanker

Authors: Associate Professor K.V.S.S.Rama Krishna, Jakka Venkata Lahari, Gurram Yasaswini, Marri Lakshmi Poojitha, Changa Nagalakshmi, Udayagiri Bhavani

Abstract: IoT-Rain Sense is an innovative and state-of-the-art solution for rain prediction on demand and continual weather monitoring based-on Internet of things (IOT) systems and cloud-based Neural Networks designed to predict precise, hyper localised forecasts. The architecture of the system consists of three main components: Data Acquisition, Feature Processing, and Weather Prediction. In phase 1, sensors being IoT based and ESP32 microcontrollers keep on monitoring temperature, humidity and light intensity over the environment of an application. The measurements are displayed in real time on a built local LCD interface. These sensors are cheap and energy-friendly, which means they could be sprinkled around agriculture and cities and institutions, without bothering anyone, and can scale up as needed. The second level is focused on feature processing, including preprocessing which aims to clean, filter and normalize raw data in order to control the quality of them. There is more weather related information added to

DOI: https://doi.org/10.5281/zenodo.20646110

Published by: vikaspatanker

Authors: Mamta Bhardwaj

Abstract: Polycystic Ovary Syndrome (PCOS) is a multifactorial endocrine disorder affecting a significant proportion of women of reproductive age, often leading to metabolic, hormonal, and reproductive complications such as infertility, insulin resistance, and cardiovascular risks. Early and accurate diagnosis of PCOS remains a major clinical challenge due to its heterogeneous symptoms, variability across patients, and reliance on subjective diagnostic criteria such as the Rotterdam guidelines. In recent years, machine learning (ML) techniques have shown promising potential in improving diagnostic accuracy; however, their lack of interpretability has limited their adoption in real-world healthcare settings. This study proposes a comprehensive Explainable Artificial Intelligence (XAI)-based risk prediction framework for PCOS diagnosis that combines robust machine learning algorithms with interpretable techniques to enhance clinical trust and usability. The proposed model utilizes a publicly available PCOS dataset comprising clinical, hormonal, and ultrasound features. A systematic preprocessing pipeline is implemented, including missing value imputation, feature scaling, and class imbalance handling using Synthetic Minority Oversampling Technique (SMOTE). Feature selection methods such as correlation analysis and Recursive Feature Elimination (RFE) are applied to identify the most significant predictors contributing to PCOS. Multiple machine learning models, including Logistic Regression, Decision Tree, Random Forest, Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost), are evaluated. A stacking ensemble model is then developed to leverage the strengths of individual classifiers and improve overall predictive performance. To address the critical challenge of model interpretability, ex-plainability techniques such as SHapley Additive explanations (SHAP) and Local Interpretable Model-Agnostic Explanations (LIME) are integrated into the framework. These methods provide both global and local explanations, enabling the identification of key features such as menstrual cycle irregularity, Body Mass Index (BMI), follicle count, and hormonal imbalance, which are consistent with established clinical knowledge.

DOI: http://doi.org/10.5281/zenodo.20643863

Published by: vikaspatanker