Performance Evaluation and Comparison of Ensemble Based Bagging and Boosting Machine Learning Methods for Automated Early Prediction of Myocardial Infarction

Myocardial Infarction occurs due to the destruction of heart tissue resulting from the obstruction of the blood supply to the heart muscle. It refers to a heart attack which is a major heart disease throughout the world. Machine learning techniques can be engaged as a decision support system for predicting myocardial infarction from a group of important predictive features that may categorize the severe-risk patients and can provide guidance to minimize the severity. In this research, we have collected myocardial infarction patient's data to assess the classification performance of two different ensemble based machine learning methods Bagging and Boosting with five different base classifiers such as Support Vector Machine, K-Nearest Neighbor, Naïve Bayes, Decision Tree, and Random Forest for predicting myocardial infarction in an earlier stage. It should be understood that finding important attributes can help to increase performance. Experimental result showed that the Bagging with Random Forest ensemble method outperformed other methods by achieving higher accuracy of 96.50%.