Neural Network-Based Advanced Cancer Prediction and Classification for Enhanced Diagnosis and Prognosis Accuracy
Authors:- Valarmathi P, Rubadharshini A K, Subashini P, Arullakshmi A
Abstract- One of the main areas of contemporary machine learning and data mining research is medical diagnostics. Since single nucleotide polymorphisms (SNPs) contribute significantly to the variability of the human genome, they have been linked to a number of illnesses, including cancer. The most prevalent malignant growth in women, breast cancer, has become much more prevalent during the last 20 years. Several methods have been used on Genetic data to make distinctions between these tumorous and benign data. The large amount of features in SNP data, which makes classification difficult, is one of the main issues.The dimensionality problem for the diagnosis of cancer in women is addressed in this research by an innovative blended intelligence technique based on Association Rules for Harvesting (ARM) and neural network technology (NN) who employs the Evolutionary Computation (EA). While NN is employed to achieve successful classification, ARM optimized by Grammatical Evolution (GE) is used to obtain relationships between SNPs, diminish dimension, which and find the most useful features. The NCBI GEO (Gene Expression Omnibus) website’s carcinoma SNP dataset was used to test the suggested NN-GEARM technique. Up to 90% consistency has been achieved by the developed model.
