IMPLEMENTATION OF THE RANDOM FOREST ALGORITHM FOR CHRONIC KIDNEY DISEASE CLASSIFICATION

Abstract

Chronic Kidney Disease (CKD) is a serious health problem characterized by a gradual decline in kidney function and often detected at an advanced stage due to the absence of early clinical symptoms. Early detection is therefore essential to reduce complications and improve patient outcomes. This study aims to implement the Random Forest algorithm for the classification of chronic kidney disease based on clinical data. The dataset used consists of 400 patient records with 26 clinical attributes, including blood pressure, creatinine, haemoglobin, sodium, potassium, albumin, and other medical indicators. Prior to model development, several preprocessing steps were performed, including handling missing values using median and mode imputation, encoding categorical variables, data normalization using Min-Max Scaling, and class balancing using the Synthetic Minority Over-sampling Technique (SMOTE). The dataset was then divided into training data (80%) and testing data (20%). The Random Forest classifier was implemented using 100 decision trees with the Gini criterion. The evaluation results show that the proposed model achieved high performance with an accuracy of 0.975, precision of 0.97, recall of 0.98, and an F1-score of 0.975. The confusion matrix analysis indicates that the model can effectively classify CKD and non-CKD cases with minimal classification errors. In addition, feature importance analysis reveals that clinical parameters such as creatinine, haemoglobin, albumin, diabetes, and hypertension play significant roles in predicting chronic kidney disease. These findings demonstrate that the Random Forest algorithm has strong potential to be utilized as a decision-support tool for early detection of chronic kidney disease based on clinical data.