Early stage Chronic Kidney Disease Prediction using Random Forest Algorithm; A machine learning approach

Kidneys are the most important organs of human body. The abnormal functionality of kidneys is referred as chronic kidney disease (CKD), which is increasing day by day and becoming a major global health issue. Early detection of CKD is essential to avoid any further loss. Machine Learning techniques playing a significant role in the early detection of disease. This research paper explores the use of the Random Forest algorithm, offering best results in the early detection of CKD. Dataset comprising of 400 patients data including 25 attributes is used which is taken from the online machine learning repository. Effectiveness of the model for detecting CKD is checked by different metrics including accuracy, F1-scrore, precision and recall. This shows 100% accurate results. Several plots are plotted including feature importance plot, confusion matrix plot, receiver operating curve plot and learning curve plot.