Chronic Kidney Disease Prediction Using Machine Learning

Abstract

In today's era everyone is trying to be conscious about health although due to workload and busy schedule one gives attention to health when it shows any symptoms of some kind. However, CKD is a disease that either exhibits no symptoms at all or exhibits no signs that are particular to the condition, making it difficult to forecast, identify, and prevent such a disease and this could be led to permanently health damage, but machine learning can be hope in this problem it is best in prediction and analysis. We will employ several machine learning approaches, such as Decision Tree, KNN, Random Forest, SVM, Naive Bayes, using data that contains 24 health related attributes like age, blood pressure, sugar, glucose, taken in 2-month period of 400 patients in which 11 numeric and 14 nominal attributes in which it consists of class label named ‘Class’ which classifies patients having disease and not present. To build a model with maximum accuracy of predicting whether CKD or not and if yes then its Severity.