A Novel Approach to Predict Cardiovascular Disease with Extra Trees Classifiers and Particle Swarm Optimization

Research indicates a progressive rise in mortality rates associated with Cardio-vascular disease. The primary causes of heart diseases and reduced blood flow through vessels stem from stress and unhealthy lifestyles. The World Health Organization suggests that current mortality rates are elevated due to widespread gadget addiction and persistent stress in daily life. Machine learning-based prediction systems have been proposed for cardiovascular diseases, leveraging their ability to uncover hidden features and complex relationships. However, these systems often lack precision in identifying the exact cause of the observed outcomes. Specifically, the study aims to enhance the predictive accuracy and robustness of the Extra Trees Classifier by employing novel data preprocessing techniques. Preprocessing helps in removing irrelevant and noisy data, thereby improving the signal-to-noise ratio and enhancing the classifier’s performance. By preprocessing the data, it becomes feasible to extract and select relevant features, facilitating better discrimination between classes and boosting the classifier’s predictive power. To improve features clarity the Novel Extra Trees Classifier (NETC) techniques perform advanced feature selection and data transformation approaches exploring the invisibility, aiming to further improve its predictive accuracy of $\mathbf{9 2. 7 \%}$ and generalization capabilities. When combined with other techniques, evaluate PSO optimization’s effectiveness in enhancing cardiovascular disease prediction models, including accuracy and ROC curve analysis.