Comparison of Classification Techniques for Heart Health Analysis System

Abstract

Heart disease diagnosis is a difficult task which requires utmost accuracy. This accuracy is achieved through knowledge and experience in the field of medicine. This paper describes a heart diagnostic system which analyses several health parameters and medical test results to predict absence or presence of heart disease in terms of artery narrowing in the patient. The proposed system described in the paper is a computer based application which uses the patient information related to the various health parameters which govern the procedure of diagnosis for a heart disease. The system also relies on pre-processing of data. The system described in this paper has been created with a view to assist doctors and medical staff in diagnosis of heart related problems. The application also facilitates self-diagnosis for the common man. This paper deals with the internal functioning of the system which is based on the data mining techniques for classification. Few algorithms for classification based mining and association based mining of data and their comparisons have been incorporated in the paper. Classification is a data mining method used to classify data into pre-defined class labels. For instance, classification can be used to anticipate the weather on a specific day . Famous grouping procedures incorporate decision trees and neural systems.