Chronic kidney disease prediction using machine learning techniques

Abstract Goal three of the UN’s Sustainable Development is good health and well-being where it clearly emphasized that non-communicable diseases emerging challenge. One objectives to reduce premature mortality from disease by third in 2030. Chronic kidney (CKD) among significant contributor morbidity can affected 10–15% global population. Early accurate detection stages CKD believed be vital minimize impacts patient’s complications such as hypertension, anemia (low blood count), mineral bone disorder, poor nutritional health, acid base abnormalities, neurological with timely intervention through appropriate medications. Various researches have been carried out using machine learning techniques on at stage. Their focus was not mainly specific prediction. In this study, both binary multi classification for stage prediction out. The models used include Random Forest (RF), Support Vector Machine (SVM) Decision Tree (DT). Analysis variance recursive feature elimination cross validation applied selection. Evaluation done tenfold cross-validation. results experiments indicated RF based has better performance than SVM DT.