Detection of Chronic Kidney Disease by Using Adaboost Ensemble Learning Approach

Abstract

Chronic kidney disease can be detected with several automatic diagnosis systems. In this study, chronic kidney diseases are diagnosed with Adaboost ensemble learning algorithm. Decision tree based classifiers are used in the diagnosis. The classification performance are evaluated with kappa, mean absolute error (MAE), root mean squared error (RMSE) and area under curve (AUC) criterias. Considering the performance analyses, it is observed that adaboost ensemble learning algorithm provides better classification performance than individual classification.