Evaluation of ensemble methods for diagnosing of valvular heart disease

In this work, we investigate the use of ensemble learning for improving classifiers which is one of the important directions in the current research of machine learning, in which bagging, boosting and random subspace are three powerful and popular representatives. Researchers have so far shown the efficacies of ensemble methods in many practical classification problems. However, for valvular heart disease detection, there are almost no studies investigating their feasibilities. Thus, in this study, we evaluate the performance of three popular ensemble methods for the diagnosis of the valvular heart disorders. To evaluate the performance of investigated ensemble methodology, a comparative study is realized by using a data set containing 215 samples. Moreover, to achieve a comprehensive comparison, we consider the previous results reported by earlier methods (Çomak, Arslan, & Turkoglu, 2007; Sengur, 2008a,b; Sengur & Turkoglu, 2008; Turkoglu, Arslan, & Ilkay, 2002, 2003; Uguz, Arslan, & Turkoglu, 2007). Experimental results suggest the feasibilities of ensemble classification methods, and we also derive some valuable conclusions on the performance of ensemble methods for valvular heart disease detection.