CLASSIFICATION OF KIDNEY CANCER DATA USING DEPTH AWARE GENERATIVE ADVERSARIAL NETWORKS APPROACH
Advanced Biotechnology methods have led the generation to Large-Scale Bioinformatics and Gene Data and makes it important to analyze this data in Bioinformatics. This study analyses Gene Expression Data from 1157 kidney cancer patients to identify particular genes for prognosis. To overcome data instability, an end-to-end, depth-aware generative adversarial networks (DAGAN) approach including a loss function for the tasks of classification is proposed. The proposed approach combines the empirical wavelet transform (EWT) to rebuild the loss in non-linear Feature Extraction and neutral network for neutral categorization loss. Medical information and genome data are utilized to define the optimum classification method and to analyze the accuracy of classification through sample category, primary detection, tumor level, vital stage as risk factors. The result of this examination shows that the DAGAN is very effectual than the typical machine learning and the data mining strategies to predict kidney cancer prognosis using gene expression data. These findings have important implications for feature extraction from gene biomarkers to predict, prevent and early detect kidney cancer prognosis.
Genomic Data, Deep Learning; Kidney Cancer; Bioinformatics