Optimizing Cardiovascular Disease Detection Using Ranking-Based Feature Selection Machine Learning Models

Cardiovascular Disease (CVD) continues to be one of the leading causes of mortality worldwide, emphasizing the urgent need for accurate and efficient diagnostic solutions. This study presents a machine learning-based framework for the classification of CVD that is transparent, reliable, and computationally optimized. The performance of Machine Learning (ML) models can be significantly hindered by class imbalances and high-dimensional datasets. To address these challenges, various Feature Selection (FS) techniques were applied to identify the most informative predictors, thereby reducing both dimensionality and computational complexity. A comprehensive evaluation of multiple ML algorithms was conducted in conjunction with FS methods, using standard performance metrics such as accuracy, precision, recall, and F1-score. The results demonstrate that integrating effective feature selection techniques not only improves model interpretability but also mitigates the risk of overfitting. The proposed system incorporates a novel feature ranking algorithm that enhances the selection of optimal predictors, playing a crucial role in the construction of robust and reliable diagnostic models. The proposed ensemble model achieved a peak accuracy of 98.20% with an 80:10:10 split for training, testing, and validation. These findings highlight the potential of the proposed model to support early clinical decision-making, enabling timely interventions and reducing the likelihood of severe CVD-related outcomes. The integration of such intelligent systems into clinical workflows may contribute significantly to improving patient care and disease management.