Multivariate discretization by recursive supervised bipartition of graph

Abstract

In supervised learning, discretization of the continuous explanatory attributes enhances the accuracy of decision tree induction algorithms and naive Bayes classifier. Many discretization methods have been developped, leading to precise and comprehensible evaluations of the amount of information contained in one single attribute with respect to the target one. In this paper, we discuss the multivariate notion of neighborhood, extending the univariate notion of interval. We propose an evaluation criterion of bipartitions, which is based on the Minimum Description Length (MDL) principle [1], and apply it recursively. The resulting discretization method is thus able to exploit correlations between continuous attributes. Its accuracy and robustness are evaluated on real and synthetic data sets. © Springer-Verlag Berlin Heidelberg 2005.