D.-S. Huang et al. (Eds.): ICIC 2012, CCIS 304, pp. 387–393, 2012. © Springer-Verlag Berlin Heidelberg 2012 A New Approach for Bayesian Classifier Learning Structure via K2 Algorithm Heni Bouhamed 1 , Afif Masmoudi 2 , Thierry Lecroq 1 , and Ahmed Rebaï 3 1 University of Rouen, LITIS EA 4108, 1 rue Thomas Becket, 76821 Mont-Saint-Aignan cedex, France 2 Department of Mathematics, Faculty of Science of Sfax, Soukra B.P 802 Sfax, Tunisia 3 Bioinformatics Unit, Centre of Biotechnologie of Sfax, 3018 Sfax, Tunisia Heni.bouhamed@yahoo.fr, Thierry.lecroq@univ-rouen.fr, Afif.masmoudi@fss.rnu.tn, Ahmed.rebai@cbs.rnrt.tn Abstract. It is a well-known fact that the Bayesian Networks’ (BNs) use as classifiers in different fields of application has recently witnessed a noticeable growth. Yet, the Naïve Bayes’ application, and even the augmented Naïve Bayes’, to classifier-structure learning, has been vulnerable to certain limits, which explains the practitioners’ resort to other more sophisticated types of algorithms. Consequently, the use of such algorithms has paved the way for raising the problem of super-exponential increase in computational complexity of the Bayesian classifier learning structure, with the increasing number of descriptive variables. In this context, the present work’s major objective lies in setting up a further solution whereby a remedy can be conceived for the intricate algorithmic complexity imposed during the learning of Bayesian classifiers’ structure with the use of sophisticated algorithms. Noteworthy, the present paper’s framework is organized as follows. We start, in the first place, by to propose a novel approach designed to reduce the algorithmic complexity without engendering any loss of information when learning the structure of a Bayesian classifier. We, then, go on to test our approach on a car diagnosis and a Lymphography diagnosis databases. Ultimately, an exposition of our conducted work’s interests will be a closing step to this work. Keywords: Bayesian Classifier, structure learning, classification, clustering, modeling, algorithmic complexity, K2 algorithm. 1 Introduction It is worth noting that efficient classifiers can be reached through the use of Bayesian networks [1, 2, 3]. In fact, a Bayesian Classifier relative to a problem with p variables is characterized by the distinction of having p + 1 nodes. Indeed, all Bayesian classifiers model the fact of belonging to a certain class by means of a discrete node dubbed "class node". This node is discrete and multinomial having k modality. The class node is distinct for not owning a parent node. Regarding the other p variables, which we call descriptive variables, they are denoted X i (i from 1 to p). The Bayesian classifier with the simplest structure is the Naïve Bayesian Network (RBN) [9], also