Implementation of the K-Nearest Neighbor Algorithm for Detecting Heart Attack Disease

Abstract

Heart damage can be fatal to human health, and even some heart disorders can cause death. A sudden reduction in blood supply to the heart can occur when one of the coronary arteries is temporarily blocked due to a blood clot. The portion of the heart muscle normally supplied by the blocked pulse ceases to function properly as soon as the plasma subsides on its own, symptoms disappear completely, and the heart muscle functions normally again. Basically, heart attack disease can be analyzed as its emergence early on. However, due to a lack of knowledge, many people are late to realize it. Early detection of heart attack can be done by analysis. In this study, the analysis method used was Machine Learning using K-Nearest Neighbor algorithm. The algorithm looks for the best results so that the initial diagnosis error can be minimized. The K-Nearest Neighbor algorithm in this study provides the best accuracy of 90.16% in classifying 303 datasets of patients with heart attack detection.