Distribution Based Chicken Swarm Optimization And Enhanced Support Vector Machine Algorithm For Chronic Kidney Disease Classification

Abstract

One of the major issues with death rates in healthcare today is Chronic Kidney Disease (CKD), that is slow to detect and often goes undiagnosed. Because of this important problem, millions of men and women suffer every year as a result of inadequate early screening programs and treatment. Nonetheless, prompt identification of the illness at its initial stages can save people's lives. Furthermore, given a trustworthy dataset, the machine learning (ML) method's assessment procedure can identify the stage of this fatal illness much more quickly. In this work, Distribution based Chicken Swarm Optimization (DCSO) algorithm and Enhanced Support Vector Machine (ESVM) algorithm is suggested. Initially, the dataset is collected which is pre-processed utilizing K-Means Clustering (KMC) system. It is utilized to handle the missing values and error rates efficiently. Formerly, the feature selection is completed by DCSO process to select the more relevant and important disease features from the pre-processed dataset. It is done based on features including age, sugar level, haemoglobin through best fitness values. After that Weighted Fuzzy C Means clustering (WFCM) process is applied to predict the data sample’sclass label to reduce the misclassification results. Finally, the CKD medical dataset classification is performed using ESVM algorithm. It performs training and testing process via weight based support vectors which improves the CKD classification accuracy significantly. According to the study findings, the suggested DCSO-ESVM method outperforms the current methods according to of greater accuracy, precision, recall, and f-measure.