Predictive Precision: An In-Depth Examination Of Machine Learning Approaches For Heart Attack Prognosis

Abstract

Currently several diseases have become epidemic in Bangladesh, one of them is heart attack. Anyone can suffer a heart attack at any age. Generally, older people and men are more prone to it. However, women are also at increased risk of heart attack as they age. Individuals who smoke, have diabetes, high blood pressure, high cholesterol, or both are also more vulnerable. A family history of coronary artery disease or ischemic heart disease increases the risk for others in the family. By applying machine learning and deep learning, this paper proposes to automate heart attack diagnosis. This dataset consists of both heart attack patients and non-heart attack patients. Heart attacks were detected and its types determined in this study, taking the classification process a step further, since most previous studies only detected heart attacks or classified them into a few types. We can solve this by using artificial intelligence, for example, With the use of machine learning and deep learning algorithms, we are able to identify heart attacks and get alerts about them. In this paper I used machine learning and deep learning algorithms, through which we used 1 model of deep learning and 4 models of machine learning. CNN model of deep learning is used here, and Random Forest, KNN, Decision Tree, SVM model of machine learning is used. The accuracy of 76% using CNN model of deep learning, and accuracy of 94.9% using Random Forest model of machine learning, accuracy of 87.6% using KNN model, accuracy of 91% using Decision Tree model 63% using SVM model. All techniques demonstrate that the Random Forest model yields the best accuracy. Using the SVM, the lowest accuracy was attained.