APPLICATION OF DECISION TREE CLASSIFIERS IN DIAGNOSING HEART DISEASE USING DEMOGRAPHIC DATA

The increase in sudden deaths due to heart disease is overwhelming. The decrease in the number of expert doctors has lead to this problem to a large extent. In recent years, machine learning methods are widely used for predictive analysis in Medical diagnosis. This paper proposes a medical decision support system for heart disease risk classification using Decision Tree Classifier. Standard Benchmark database of heart disease from University of California, Irvine (UCI), literature from journals and expert discussions have contributed in designing the attribute set for Ischemic Heart disease for a demographic population in India. The data set is collected from cardio thoracic department, Madras Medical College, Chennai, India from January to April 2011. A total of 712 patients examined are classified into 4 classes using C 4.5 Decision Tree Classifier. Pre-processing steps filtered the data set to 662 patients. The accuracy of correctly classified instances is 82.33%. The same data set with 2 class output gives an accuracy of 92.56%. Whereas the accuracy of UCI Cleveland data set is 78.91%. . The main aim of is to develop more cost-effective and easy-to-use systems, procedures and methods for supporting clinicians