Medical data classification with Naive Bayes approach

Abstract

Medical area produces increasingly voluminous amounts of electronic data which are becoming more complicated. The produced medical data have certain characteristics that make their analysis very challenging and attractive. In this study we present an overview of medical data mining from different perspectives; including characteristics of medical data, requirements of systems dealing with such data and the different techniques used for medical data mining. Among the different approaches we emphasize on the use of Naive Bayes (NB) which is one of the most effective and efficient classification algorithms and has been successfully applied to many medical problems. To support our argument, empirical comparison of NB versus five popular classifiers on 15 medical data sets, shows that NB is well suited for medical application and has high performance in most of the examined medical problems. © 2012 Asian Network for Scientific Information.