Performance Investigation of Different Machine Learning Algorithms in Predicting Chronic Kidney Disease

Abstract

This paper implies an investigative approach of studying the performance of different boosting algorithm in predicting chronic kidney diseases more accurately. In recent years chronic kidney disease (CKD) has reached a global prevalence as high as 11–13% with the majority in stage 3 which can lead to end stage renal disease (ESRD) if not detected early. Different boosting machine learning algorithms has been proven to be an effective tool to detect CKD while it’s still in one of its initial stages. A dataset containing 400 instances and 25 attributes from the University of California, Irvine (UCI) repository has been exploited to train and test the model classifier. Four different data frames and correlation heatmap were constructed by four different strategies to begin the operation of the classifiers. Eleven machine learning algorithms were studied and their performance parameters like confusion matrix and accuracy were analyzed. Furthermore, a broad comparative investigation was conducted through the simulation of precision, sensitivity, F1 score, ROC-AUC of each algorithm.