Severity Prediction and Multi Classification of Chronic Kidney Disease Based on Machine Learning Techniques

Abstract

One of the major causes of morbidity and mortality from non-communicable diseases is chronic kidney disease (CKD). Symptoms do not appear till kidneys lose most of its functionality. Hence, early and precise CKD stage detection can minimize the impact on the health of patients. Moreover, Further complications such as hypertension, anemia, brittle bones and nerve damage can be reduced. Recently, machine learning techniques are widely employed for the prediction and classification of diseases in healthcare system. This study focuses on the use of machine learning techniques for specific stage prediction and detection of CKD. The proposed model involves applying a set of seven distinct ML based classification models such as Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), Naïve Bayes (NB), Extreme Gradient Boosting (XGBoost), Adaptive Boosting (AdaBoost). Several experiments were conducted in this study including different data imputation techniques and feature selection methods. The assessment of these models has done based on four performance metrics including accuracy, precision, f-measure, and recall. Results had indicated that XGBoost and RF outperformed other techniques.