Performance Evaluation of Ensemble-based Machine Learning Techniques for Prediction of Chronic Kidney Disease

Abstract

Chronic kidney disease (CKD) is widespread and related with enhanced risk of cardiovascular disease and end-stage renal disease, which are possibly escapable through early detection and treatment of individuals at risk. Machine learning algorithm helps medical experts to diagnose the disease correctly in the earlier stage. Therefore, machine-predicted analysis has become very popular in recent decades that can efficiently recognize whether a patient has certain kidney disease or not. In this regard, we propose an ensemble method based classifier to improve the decision of the classifiers for kidney disease diagnosis efficiently. Ensemble methods combine multiple learning algorithms to achieve better predictive performance than could be obtained from any of the constituent learning algorithms. In addition, Data is evaluated by using tenfold cross-validation and performance of the system is assessed on receiver operative characteristic curve. Extensive experiments on CKD datasets from the UCI machine learning repository show that our ensemble-based model achieves the state-of-the-art performance.