Forecasting of Chronic Kidney Disease and Analysis of the Classifiers using ML based Classification Approaches

Abstract

The rising number of deaths due to kidney failure is of great concern. Chronic Kidney Disease (CKD) is a disorder where people’s kidneys are damaged and unable to filter blood properly. This damage can cause accumulation of toxins in the body. CKD will become irreversible if timely treatment is not given. Many strategies and machine learning techniques have been developed to detect CKD at an initial state of the disease. Out of all available strategies, machine learning (ML) techniques have been contributing a major role in the early prediction of several diseases. The ML methods used to obtain analytical results have shown great performance in terms of analysis and prediction. In this research paper, we use predictive analysis of ML classifiers for categorizing the patient’s kidney data either as Chronic Kidney Disease or No Chronic Kidney Disease. Here the algorithms namely, Support Vector Machine (SVM), K-Nearest Neighbor (KNN) and Random Forest (RF) have been used in the prediction of the disease. The performances were evaluated using various accuracy criteria.