Performances Analysis of Heart Disease Dataset using Different Data Mining Classifications

Abstract

nowadays, heart disease is one of the major diseases that cause death. It is a matter for us to concern in today’s highly chaotic life style that leads to various diseases. Early prediction of identification to heart-related diseases has been investigated by many researchers. The death rate can be further brought down if we can predict or identify the heart disease earlier. There are many studies that explore the different classification algorithms for classification and prediction of heart disease. This research studied the prediction of heart disease by using five different techniques in WEKA tools by using the input attributes of the dataset. This research used 13 attributes, such as sex, blood pressure, cholesterol and other medical terms to detect the likelihood of a patient getting heart disease. The classification techniques, namely J48, Decision Stump, Random Forest, Sequential Minimal Optimization (SMO), and Multilayer Perceptron used to analyze the heart disease. Performance measurement for this study are the accuracy of correct classification, mean absolute error and kappa statistics of the classifier. The result shows that Multilayer Perceptron Neural Networks is the most suited for early prediction of heart diseases.