A LIME-Based Explainable Machine Learning Technique for the Risk Prediction of Chronic Kidney Disease

Chronic Kidney Disease (CKD) is on the rise in the current research scenario. The main function of the kidney is to remove and purify the waste and blood in the human body. Diabetes is the most prevalent cause of kidney disease. The key to preventing or curing CKD is identifying it at an early stage. If early detection is avoided, there is a greater chance of kidney failure as well as heart disease, bone disease, or an imbalance in potassium and calcium levels. Prediction at an early stage for a long and healthy life is made feasible with the assistance of a machine intelligence classifier. In this study, the prediction of CKD and Non-CKD patients is done by applying five machine learning classifiers. The results show that the Random Forest classifier has the greatest accuracy of all classifiers. Explainable artificial intelligence (XAI) is introduced for a trustworthy explanation of the result. XAI investigates how the Random Forest model gives high accuracy with input features imported into the classifier.