A Majority Hard Voting Based Ensemble Approach to Predict Cardiovascular Disease

Cardiovascular Disease (CVD) remains a major global health concern that continues to cause significant disability and death. Early and precise diagnosis of cardiac disease is essential to prevent more harm and prolong patients’ lives. Resolving this acute issue requires novel approaches for early identification and risk evaluation. Machine learning has captivated the attention of scholars in the medical sciences. Researchers employ diverse machine learning (ML) techniques and methodologies for predicting cardiac disease. This study investigates the application of ML in the prediction of heart disease. The main goal is to improve the accuracy of diagnoses and make it easier for healthcare professionals to take preventative steps. We conducted a comparative analysis using widely used models such as Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), Logistic Regression (LR), and Extreme Gradient Boosting (XGBoost). Of these five ML models, XGBoost and Random Forest attained a remarkable accuracy of 94.95%. Using a majority hard voting-based ensemble approach, we obtained an amazing accuracy of 95.79%, therefore supporting the general effectiveness of predictive modeling. The outcomes of this research will contribute significantly to machine learning (ML)based prediction of heart disease and its associated ailments.