Heart Disease Prediction

Abstract

Heart disease remains the leading killer in the world. Nearly 80% of deaths occurred in low- and middle-income nations. If current trends continue, approximately 23.6 million people will die from cardiovascular disease (primarily heart attacks and strokes) by 2030. The healthcare industry gathers massive amounts of data on heart disease, which is generally not "mined" for hidden information that will help decision- makers make more informed decisions. Heart disease is caused by a decline in blood and oxygen delivery to the heart. However, reliable analysis methods for discovering hidden linkages and trends in data are lacking. This proposed study aims to present a survey of current knowledge discovery strategies in databases utilizing Decision tree classifiers, which will be valuable for medical practitioners in making good decisions. The goal of this study is to find a way to predict the existence of heart disease with a smaller number of variables. Researchers use a variety of machine learning approaches to assess huge amounts of complex medical data, assisting doctors in the prediction of cardiac disease. There are 303 instances and 76 attributes in the collection. Only 14 of the 76 attributes are considered for testing, which is essential for evaluating the performance of different algorithms. We have used Logistic Regression algorithm to develop a user-friendly way to predict heart disease and deployed it in a web application. KEYWORDS: Machine learning, Decision tree, Heart disease prediction, Flask