Comparison of Machine Learning Models for Chronic Kidney Disease Prediction using UCI dataset

Chronic Kidney Disease (CKD) prediction remains a formidable obstacle in the field of medicine. CKD is a potentially fatal condition that may continue for the rest of a person’s life and is commonly associated with kidney cancer or impaired kidney function. The development of CKD to its terminal stage can be prevented or reduced by means of timely identification and suitable treatment. The present study investigates the potential of different Machine Learning (ML) predictive models to aid in the early detection of CKD. Utilizing publicly accessible data to predict the occurrence of CKD is the main goal of this study. In order to accomplish this, a generic model was constructed through the implementation of a series of data pre-processing steps. These steps encompass tasks such as missing data handling and encoding, in addition to the application of the Chi-Squared (Chi-2) algorithm for feature selection. ML techniques including Support Vector Machine (SVM), Decision Tree (DT), and Convolutional Neural Network (CNN) were employed to develop predictive models. Among these, the CNN model emerged as the most promising, achieving the highest number of correct predictions with 119 cases, while making only one incorrect prediction.