Date of Publication :2nd September 2024

Abstract:This study addresses the urgent global health problem of heart disease (HD) by proposing an ensemble machine learning architecture for cardiovascular disease prediction. Typical HD symptoms include frailty, difficulty of breath, and swollen feet, but traditional diagnostic methods lack efficiency and precision. Accuracy is improved by using feature selection techniques like Recursive Feature Elimination (RFE) and Principal Component Analysis (PCA) inside the suggested model, which is an amalgamation of the Elastic Net, Logistic Regression, Gradient Boosting, and Extreme Gradient Boosting (XG Boost) algorithms. It makes use of all obtainable features for prediction using parallel processing. Compared to state-of-the-art approaches, experimental data show that the diagnosis accuracy is significantly higher. The precision of 92.86%, recall of 85.22%, F1-Score of 88.87%, and total accuracy of 88.30% are just some of the impressive metrics demonstrated by the models in the present research. In addition, the robustness of the model is validated by an Area Under the Curve (AUC) value of 0.95. The suggested ensemble model demonstrates superiority in precision when compared to other methods, which makes it a potentially useful technique for the prediction of cardiovascular disease.

Reference :