Enhanced Cardiovascular Disease Prediction: A Comparative Analysis of Machine Learning Models with SMOTE Balancing

Cardiovascular disease (CVD) is one of the leading causes of mortality worldwide, accounting for over 32% of deaths in several countries worldwide. Early detection of CVD is crucial for saving lives, as traditional diagnostic methods can be time-consuming and prone to inaccuracies. This study aims to evaluate various machine learning (ML) models for the accurate prediction of CVD. The models assessed include Extra Trees Classifier, AdaBoost, Random Forest, XGBoost, Light GBM, Gradient Boosting, and Stacking Classifier. The dataset is collected from Kaggle and preprocessed, followed by splitting it into training and testing subsets. Different types of data preprocessing techniques, including feature transformation, normalization, the Synthetic Minority Over-sampling Technique (SMOTE) for class balancing, and encoding, were applied to ensure a clean and well-structured dataset for training the model. Extensive experiments were conducted utilizing the CVD dataset, with all models trained and evaluated simultaneously for fair evaluation. Among these, the Stacking Classifier achieved the highest accuracy of 90.17%. Additionally, Random Forest (RF), LGBM Classifier, and XGBoost demonstrated strong performance, achieving accuracies of 85.68%, 83.70%, and 87.70%, respectively.