Comparative Analysis of Accuracy on Heart Disease Prediction using Classification Methods

Abstract

A common term heart disease is nothing but a cardiovascular disease or a Coronary heart disease which reduces the efficiency and proper functioning of heart by blocking veins, artery or blood vessels around it. Coronary heart disease causes disability such as damage to the brain resulting in death. Based on Statistics [10] it indicates that range of age group from 25 to 69 have 25% risk of having heart diseases. Some vital causes for cardiovascular disease are, physical inactivity, smoking, consuming more junk food and addiction of alcohol which are major causes for stroke, chest pain, and heart attack. However because of the awareness about factors and symptoms that are responsible for heart problem, it is possible to predict any heart problem based on statistical analysis of medical records. However Data mining, a modern technique has provided an automatic way of analyzing data using standard classification methods. Though many classifiers are available in data mining that can be used to predict the heart problems, this paper emphasizes on finding the appropriate classifier that has the potential to give better accuracy by applying data mining techniques viz. Naïve Bayes , Support Vector machine and Logistic Regression.