Chronic Kidney Disease (Ckd) Prediction by Supervised Machine Learning Techniques

The function of kidney is filtering waste and excess fluid from our body which is then let out through urine excretion. Due to Chronic Kidney disease there will be kidney impairment. This may also lead to several complications like anaemia, high blood pressure, anorexia, and mental and physical health issues. In India, there was a 38% increase in the number of people who succumbed to kidney failure between 2001–03 and 2010–2013 which sums up to about 10 crore cases in the previous year. Thus, the prediction of this disease before end-stage renal failure is much needed. This model uses binary text classification and machine learning algorithms. We have assessed the performance of different models on CKD and the following are the results: Naive Bayes Classification as accuracy of 95.83%, XGBoost Classification as an accuracy of 100%, Logistic Regression achieved an accuracy of 99.16%, SVM Classification as an accuracy of 98.3%, Random Forest Classification as an accuracy of 100%, Artificial Neural Network (ANN) as an accuracy of 98.33%, K-Nearest Neighbors Classification (KNN) achieved an accuracy of 96.67%, Gradient Boost Classification as an accuracy of 100%, AdaBoost Classification as an accuracy of 100%, and Decision Tree Classification as an accuracy of 99.16%.