HEART ATTACK PREDICTION BASED ON THE PATIENT'S CONDITION USING MACHINE LEARNING ALGORITHMS

Abstract

A heart attack or acute myocardial infarction is the death of part of the heart muscle due to a sudden cessation of circulation through one of the arteries that feed the heart. It is among the most common causes of death in developed and developing countries. It affects men more often than women, and after entering the climax, the risk of the disease becomes even. It is more common in people over 40 years old. Cardiovascular risk factors differ; some can be influenced, and others cannot. The process of detecting or predicting a heart disease in a person can be very unpredictable, long-lasting, and often unsuccessful. Thus, applying artificial intelligence to solve this problem is becoming more common today. This paper used machine learning models to predict whether a patient will have a heart attack based on his health condition. The following models were trained: logistic regression, Naïve Bayes, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Random Forest method, Light Gradient-Boosting Machine (LightGBM), eXtreme Gradient Boosting (XGBoost) and Categorical Boosting (CatBoost). The last model was the best-performing model, and it was interpreted to uncover factors most affecting heart attack prediction.