Machine Learning Based Chronic Kidney Disease Prediction Model

Abstract

Chronic kidney disease was the most and primary cause of death internationally. If people build up CKD then their kidney become damage and over time may not clean the blood like as healthy kidney. According to WHO, most of developing countries like Ethiopia CKD is a growing problem and majority of patients with kidney disease die because lack of proper handling or give treatments. ML plays a huge role in healthcare system and it can effectively help and assist on decision support on medical centers. The main purposes of this study develop and recommend a machine learning algorithm for prediction of CKD. The CKD dataset collected from TASH which has 2135 with huge number of missing value and then after effectively preprocessing the dataset we use four types of machine learning algorithms, these are support vector machine, Decision tree, Random forest and multi- layer perceptron Artificial neural network. From those machine learning algorithms, MLP ANN the best performance with 99% accuracy.