Feature Selection and Explainable AI for Heart Disease Detection using Machine Learning

Resky Ayu Dewi Talasari
Ayutri Wahyuni (Universitas Fajar)
Clara Diva (Universitas Kristen Indonesia Paulus)
Muhammad Nur Alamsyah Rajab (Universitas Islam Negeri Alauddin)

Sistemasi: Jurnal Sistem Informasi
Vol 15, No 6 (2026): Sistemasi: Jurnal Sistem Informasi

Early detection of heart disease is essential for supporting timely clinical intervention, improving treatment outcomes, and enhancing the quality of patient care. This study compares the performance of three machine learning algorithms—Random Forest, XGBoost, and Support Vector Machine (SVM)—combined with two feature selection methods, Chi-Square and Recursive Feature Elimination (RFE), using the UCI Heart Disease dataset. Six modeling scenarios were evaluated based on accuracy, precision, recall, and F1-score. The experimental results demonstrate that the Random Forest model achieved the best overall performance, with an accuracy of 85.2% and a recall of 97.0%, indicating a strong capability to identify patients with potential heart disease. To enhance model transparency and interpretability, SHAP (SHapley Additive exPlanations) was employed as an Explainable AI (XAI) technique and integrated into a web-based decision support system to provide intuitive explanations of prediction outcomes. The proposed system is intended to serve as an initial clinical decision-support tool and is not designed to replace diagnosis or clinical judgment by healthcare professionals.