Predicting the Risk of Myocardial Infarction (MI) using Machine Learning (ML)

Abstract

The purpose of this study was to use machine learning to forecast the likelihood of a myocardial infarction and identify when one would occur. The study made use of a preprocessed dataset regarding cardiac attacks from Kaggle. The purpose of this study was to use machine learning to forecast the likelihood of a heart attack and identify when one would occur. The research used a preprocessed dataset from Kaggle related to heart attacks. K-nearest neighbors (KNN), Support Vector Classifier (SVC), Logistic Regression, Decision Tree, Naive Bayes, XGBoost, Random Forest, and Gradient Boosting were the eight techniques used in the study. According to the findings, the models that predicted heart attacks in our dataset the best were Decision Tree and Gradient Boosting. These models showed excellent precision, recall, and F1-score balance, among other important criteria. They can effectively reduce overfitting and generalize well to new data thanks to their ability to handle complicated, non-linear interactions and their use of regularization and ensemble learning techniques. Decision Tree and Gradient Boosting are the most reliable options for this predictive task because of their all-encompassing strengths, even if models like XGBoost and Random Forest also performed well while Logistic Regression and SVC produced strong results.