Ev-MDP : A Novel Metaheuristic Technique for Heart Disease Detection
Somya Goyal
Abstract
Heart Disease (HD) refers to a wide range of Cardio Vascular Diseases (CVD) that may be plaque deposits in the arteries or any other abnormality in the heart or blood vessels. Such abnormalities are very small but hard to detect. Sudden closure of one of the arteries in the heart or abrupt rupture of an artery can lead to a Heart Attack which may fall to premature death. Nowadays, an unhealthy, and stressful lifestyle has become vogue. It causes cardiovascular abnormalities. Early detection of such abnormalities allows alerting the person in advance and facilitates taking precautions. Meta-heuristic techniques are prominently being used for diagnosing heart disease by analyzing the healthcare dataset. A novel technique namely Evolution with Mathematical Diversification of Population (i.e., Ev-MDP) is proposed for early Heart Disease Detection. The datasets used are Statlog Heart Dataset from the UCI repository. The suggested technique is statistically compared with the state-of- the-art metaheuristic techniques. It is concluded that the Ev-MDP has a promising performance with values of 98.8%, 99.39%, 98.99%, 98.6%, and 98.8% for the Accuracy measure, AUC score, Precision, Recall, and F-measure respectively after conducting an empirical comparison with the state-of-art models.