← Back to Publications List

Fair and Explainable Machine Learning Framework for Heart Disease Prediction

Students & Supervisors

Student Authors
Shuvo Sarker Joy
Bachelor of Science in Computer Science & Engineering, FST
Swarup Biswas
Bachelor of Science in Computer Science & Engineering, FST
Supervisors
Prof. Dr. Kamruddin Nur
Professor, Faculty, FST
Rifat Al Mamun Rudro
Lecturer, Faculty, FST

Abstract

This study presents a comprehensive evaluation of five machine learning classifiers for heart disease detection, emphasizing not only predictive accuracy but also fairness, probability calibration, and interpretability. Using stratified 10-fold cross-validation, the models were assessed across multiple performance dimensions, with Random Forest and XGBoost achieving the best results (≈99.6% accuracy and ≈0.999 AUC). SHAP and LIME analyses identified the most influential clinical features and improved model transparency, while subgroup analysis revealed slight disparities in performance for younger patients and females. The proposed evaluation framework demonstrates that integrating accuracy with fairness, calibration, and interpretability can produce more reliable and clinically applicable heart disease prediction models.

Keywords

Disease Prediction Supervised Learning SHAP Feature Importance Model Fairness.

Publication Details

  • Type of Publication:
  • Conference Name: International Conference on Quantum Photonics, Artificial Intelligence & Networking (QPAIN)
  • Date of Conference: 16/04/2026 - 16/04/2026
  • Venue: Chattogram, Bangladesh
  • Organizer: IEEE Photonics Society Bangladesh Chapter
View on Publisher