Metric-Weighted Voting Classifier with SMOTEENN for Enhanced Machine Learning Cardiovascular Disease Prediction

This chapter introduces a comprehensive framework that combines SMOTEENN resampling with a metric-weighted voting classifier to improve the prediction of cardiovascular diseases. The framework is designed to address the limitations of traditional ensemble methods by dynamically assigning classifier weights based on clinically meaningful performance metrics such as F1-score, MCC, and ROC-AUC. The study employs seven diverse machine learning models, including Decision Trees, Logistic Regression, Random Forests, LightGBM, Multi-Layer Perceptron Classifier, k-Nearest Neighbors, and XGBoost. The framework is tested on a large-scale dataset containing 70,000 records, demonstrating substantial improvements in multiple evaluation metrics relevant to clinical decision-making. The results show that the metric-weighted voting classifier, when combined with SMOTEENN, significantly enhances the accuracy, F1-score, and ROC-AUC compared to baseline models. This approach is particularly suitable for clinical screening scenarios where reducing false negatives is crucial. The chapter also provides a detailed comparison with related work, highlighting the advantages of the proposed framework. Overall, the study presents a robust and clinically relevant method for predicting cardiovascular diseases, offering a promising tool for early intervention and preventive care.