Hybrid Fuzzy-SV Clustering for Heart Disease Identification

The identification of different heart diseases plays an important role in medical applications since it is becoming a growing problem. In order to decrease the number of deaths, it is important to consider warning signs, and knowing how to respond quickly and properly when it occurs. In this paper we propose the use of Fuzzy Support Vector Clustering in order to identify a heart disease and also to identify different degrees of sickness that serve as warning signs for patients. The algorithm uses a kernel induced metric to assign each data to a cluster and the SVM density estimation algorithm to parameterize clusters (to identify membership degrees matrix). Experimental results were performed using a well known benchmark of heart diseases.