Prediction of Chronic Kidney Disease Using Different Machine Learning Methods

Abstract

Medical science uses the term "Chronic Kidney Disease" (CKD) to refer to a set of conditions that lead to kidney damage or a low Glomerular Filtration Rate (GFR). Medical advancements in recent years have allowed doctors to apply a wide range of techniques in the treatment of this illness. Recently, AI and ML have been increasingly adopted as a useful method for improving healthcare and medical research. The use of Machine Learning to detect the early symptoms of kidney condition can be a helpful approach as the disease may lead to a life-threatening condition. Different machine learning techniques, programs, and algorithms can be applied together to predict the steady progress of Chronic Kidney Disease. An appropriate result is produced by a machine-learning algorithm, and the algorithm with the highest performance among all others is chosen as the best one. Our web based system could allow doctors to determine the formation of the disease as soon as they receive the dialysis report. Also, the report analysis can help to figure out which elements in the human body are the root cause of Chronic Kidney Disease. Complex and dynamic algorithms such as Naive Bayes, Random Forest, KNN, Decision Tree, AdaBoost & XGBoost etc. are implemented in order to achieve optimal results in this system.