AI-Driven Hybrid Decision Support Framework for Heart Disease Prediction and Personalized Treatment Recommendation

Abstract

Heart sickness remains a main cause of global mortality, yet maximum present computational fashions are confined to binary prediction and lack clinically actionable steering or interpretability. To deal with these boundaries, this paper proposes HPER – Heart (Hybrid Ensemble ML Models for Predictions with Explainability and AI-Driven Recommendations) because the middle predictive set of rules within an AI-pushed hybrid selection help framework. HPER integrates a supervised hybrid ensemble with Generative AI to provide accurate cardiovascular illness prediction and custom designed healing hints. The predictive module leverages a hybrid ensemble of XGBoost, Random Forest, Gradient Boosting, Logistic Regression, and Support Vector Classifier, making sure robust overall performance at some point of numerous affected individual profiles. Dataset imbalance is mitigated the usage of SMOTE to decorate fairness and enhance sensitivity for minority classes. A complementary Generative AI module synthesizes individualized drug, life-style, and preventive pointers primarily based on clinical information. Explainable AI (XAI) strategies are incorporated to interpret model conduct, making sure transparency in medical choice-making. The framework is verified the usage of the UCI Heart Disease dataset and a private scientific dataset, demonstrating advanced accuracy, adaptability, and actual-international applicability. By combining excessive-precision prediction with interpretable outputs and actionable pointers, HPER – Heart bridges an essential gap in current studies and supports greater knowledgeable cardiovascular care.

Ähnliche Arbeiten