Benchmarking Ensemble Learning Approaches for Coronary Artery Disease Classification

Abstract Coronary artery disease (CAD) is one of the leading causes of death worldwide, and early diagnosis plays a crucial role in reducing its impact. Traditional diagnostic methods have limitations in their ability to predict CAD accurately, which has led to the adoption of machine learning (ML) models for enhancing prediction accuracy. Among these, ensemble learning methods have shown considerable promise by combining multiple classifiers to boost the overall performance and address the complex nature of CAD prediction. Ensemble techniques such as bagging, boosting, and stacking improve model robustness and generalization. In this review, we explore various contemporary state-of-the-art algorithms used for CAD prediction, such as Random Forest, Support Vector Machines (SVM), Artificial Neural Networks (ANNs), and K-Nearest Neighbors (KNN). We compare the performance of these models with ensemble learning methods, highlighting their strengths and weaknesses in clinical settings. Additionally, the paper discusses the importance of feature selection techniques such as Principal Component Analysis (PCA) and Recursive Feature Elimination (RFE), which enhance model accuracy by eliminating irrelevant attributes. The proposed ensemble method achieved 92.5% accuracy and 0.89 AUC, outperforming traditional models by 5–7%. We also emphasize the challenges associated with ensemble models, including computational cost, model interpretability, and data imbalance, which often occurs in medical datasets. Furthermore, the review provides insights into the integration of ensemble learning models into healthcare systems and their potential for improving early CAD diagnosis and patient outcomes. Finally, the paper suggests directions for future research to improve the efficiency and applicability of ensemble methods for CAD classification, particularly in real-world clinical environments.