Efficient and scalable multi-class classification using naïve Bayes tree

Abstract

Decision tree (DT) and naïve Bayes (NB) classifiers are useful, efficient and commonly used for solving multi-class classification tasks in machine learning and data mining. In this paper, we introduce an adaptive naïve Bayes tree (NBTree) algorithm for scaling up the classification accuracy of multi-class classification problems, which considers the attributes that are used in the decision tree for the calculation of the naïve assumption of class conditional independence. The NBTree is a hybrid classifier of decision tree and naïve Bayes classifiers. In NBTree nodes contain and split as regular decision tree, but the leaves are replaced by naïve Bayes classifier. We tested the performance of proposed algorithm against those of the existing decision tree and naïve Bayes classifiers respectively using the classification accuracy, precision, sensitivity-specificity analysis, and 10-fold cross validation on 10 real benchmark datasets from UCI (University of California, Irvine) machine learning repository. The experimental results show that the proposed approach improved the classification accuracy of real life challenging multi-class problems.