Rule-Based Data Mining for Diagnosis of Coronary Heart Disease

Abstract

Coronary heart disease is the leading cause of human death due to the presence of plaque (fat) in the blood vessels. Electrocardiograph (ECG) and treadmill tests are commonly used for coronary disease detection. However, it is costly, at risk and sometimes the diagnosis result is not accurate. This research aims to classify coronary heart disease dataset based on two rules of data mining methods, i.e. variable precision rough set (VPRS) and repeated incremental pruning to produce error reduction (RIPPER). These rules are chosen to observe the simplest pattern of rules knowledge from big data, imprecise and ambiguous data. The proposed method is evaluated on Cleveland coronary heart disease dataset taken from the UCI repository. The combination of VPRS and RIPPER obtains the best evaluation result with accuracy achieved of 92.99%. While the accuracy of VPRS and RIPPER is merely 75.22% and 88.13%, respectively. It indicates that the proposed method successfully classifies coronary heart disease dataset and has a potential to be implemented in the development of a computerised coronary heart disease diagnosis system.