HSM-XAI: Hybrid Stacking Model with Explainable AI for Heart Disease Prediction
Keywords:
Heart Disease Prediction, Explainable AI, Stacking Ensemble, SMOTE, SHAP, Machine Learning, Healthcare AnalyticsAbstract
Heart disease is still the top cause of death in the world, so we need accurate and reliable prediction models to find it early and help people who need it. Machine learning (ML) has come a long way, but some problems still exist like class imbalance and results are not
interpretable by medical practitioner. Our work talks about HSM-XAI, which stands for "Hybrid Stacking Model with Explainable AI." It is a strong and easy-to-understand solution that uses hybrid stacking ensembles and SHAP-based explainability. Our method uses neural networks and adaptive boosting as base learning along with an Extra Trees classifier as a meta-model. SMOTE is used to fix dataset imbalance. The HSM-XAI is a useful algorithm for predicting heart disease at early stage because its result shows higher accuracy compare to other standalone model.
