Prediction of Myocardial Infraction Using Machine Learning Algorithms

During the supply of blood to the heart muscle is cut off, a dangerous and life-threatening condition known as a heart attack results. Heart attack early detecting and preventing can save lives and lower medical expenses. A machine learning model is offered to anticipate the likelihood of a heart attack based on a variety of medical attributes, which includes age, sex, chest pain (cp), resting of blood pressure, cholesterol level, fasting blood sugar, electrocardiography, and many more. We train and test our prediction model on a dataset of 303 patients with and without heart disease using a variety of classifier algorithms, including Random Forest using Decision Tree, K-Nearest Neighbours, Logistic Regression, Naive Bayes, Gradient Boosting Classifier, Support Vector Machine (SVM), Decision Tree Classifier, Bagging Classifier, and Ada Boosting classifier. Utilising criteria for classification report, we compare the classifiers' performance. We discover that among the classifiers, the logistic regression technique gets the greatest accuracy of 80.20% and F1-score of 0.83. We also observe the significance of the medical characteristics and their relationship to the risk of heart attack. We come to the inference that our machine learning approach can offer a trustworthy and effective tool for heart attack analysis and prediction.