Computational Intelligence Approaches for Heart Disease Detection

Abstract

A disease is an occurrence that affects one or more areas of a person's body. Various diseases are on the rise as a result of changing lifestyles and patrimonial values. Heart disease (HD) is the most serious of all disorders, and its consequences are much more dangerous than those of any other disease. Therefore, early detection of HD will reduce the death rate of people. Computational Intelligence (CI) approaches can be employed for the early diagnosis and detection of HD. Hence, employed the use of two computational intelligence approaches. The study compared a variety of computational intelligence strategies for heart disease detection. A comparison analysis was drawn using Two computational intelligence techniques: Decision Tree (DT) and K-Nearest Neighbor (KNN). A feature extraction algorithm which is Autoencoder was employed to reduce the number of attributes required to describe the heart disease dataset. The performance of each approach was measured using heart disease databases obtained from the National Health Service (NHS) database and uncertainty matrix success assessment metrics.