Enhancing Cardiovascular Disease Prediction Using Ensemble Learning

Abstract

The proposed ensemble learning framework integrates diverse machine learning algorithms. Each base model is trained on a diverse set of features derived from comprehensive patient data. To evaluate the ensemble model's performance, a large-scale dataset comprising anonymized electronic health records from a diverse patient population is employed. The dataset includes longitudinal data, allowing for the incorporation of temporal information. Preliminary results demonstrate that the ensemble learning approach outperforms individual models in terms of predictive accuracy and stability. Moreover, the model exhibits robustness across different subpopulations, indicating its potential for generalizability and applicability in diverse healthcare settings. The accuracy metrics is being utilized as a foundation for assessing the effectiveness of all classifications. With an accuracy of 93.01% the stacking classifier has proven to be the finest model.