Analisis Kinerja Random Forest untuk Prediksi Penyakit Hipertensi

Hypertension is one of the major cardiovascular diseases that contributes significantly to global mortality and disability rates and is widely recognized as a silent killer due to its frequent absence of early symptoms. The complexity of hypertension risk factors including demographic, clinical, anthropometric, lifestyle, and medical history variables necessitates a machine learning based predictive approach capable of producing accurate and consistent classifications to support early disease detection. This study develops a Random Forest model to predict hypertension risk using the Hypertension Risk Prediction dataset obtained from Kaggle. Data processing was conducted systematically through data cleaning, handling of missing values, outlier treatment using the IQR-based capping method, exploratory data analysis, feature transformation through appropriate encoding techniques, stratified train test data splitting, and numerical feature scaling. The model was subsequently evaluated using accuracy, precision, recall, F1-score, and confusion matrix metrics on the test dataset. The evaluation results indicate that the Random Forest model achieved high classification performance, with an accuracy rate of 95.47%, accompanied by balanced performance across classes. These findings suggest that Random Forest has strong potential as an interpretable and effective hypertension prediction model to support early detection efforts and the prevention of cardiovascular complications.