Technical Report No: BU-CE-1001 A Discretization Method based on Maximizing the Area Under ROC Curve

We present a new discretization method based on Area under ROC Curve (AUC) measure. Maximum Area under ROC Curve Based Discretization (MAD) is a global, static and supervised discretization method. It discretizes a continuous feature in a way that the AUC based only on that feature is to be maximized. The proposed method is compared with alternative discretization methods such as Entropy-MDLP (Minimum Description Length Principle) which is known as one of the best discretization methods, Fixed Frequency Discretization (FFD), and Proportional Discretization (PD). FFD and PD are proposed recently and designed for naïve Bayes learning. Evaluations are performed in terms of M-Measure, an AUC based metric for multi-class classification, and accuracy values obtained from naïve Bayes and Aggregating One-Dependence Estimators (AODE) algorithms by using real world datasets. Empirical results show that our method is a candidate to be a good alternative to other discretization methods.