Chronic Kidney Disease Prediction Using Random Forest, Decision Tree and Ada Boost Classifier

Millions of individuals worldwide are afflicted with the common and possibly fatal ailment known as chronic kidney disease (CKD). By allowing for prompt diagnosis and care, early identification and precise prediction of CKD may greatly improve patient outcomes. Machine learning algorithms have become effective resources for forecasting illness outcomes based on patient data in recent years. In order to better predict CKD, this research compares three well-known machine learning algorithms - Random Forest, Decision Tree, and Ada Boost Classifier performance. The clinical and laboratory data from a cohort of CKD patients were gathered to create the dataset utilized in this investigation. Demographic data, medical history, vital signs, and the findings of laboratory tests are among the characteristics. To improve the prediction accuracy of these ML algorithms, K-Fold validation techniques is applied. The findings show that Random Forest, Decision Tree, and Ada Boost Classifier might be useful tools for the early diagnosis and prediction of CKD. Among these three the prediction accuracy of Random Forest Classifier is found 99.98% which is maximum among the three.