Enhancing the performance of decision tree: A research study of dealing with unbalanced data

Computational intelligence techniques are proved to be outperforming compared to standard statistical techniques, specifically when dealing with large, unbalanced and high dimensional data. In this paper we present an enhancement approach for improving the performance of decision tree using Support Vector Machine (SVM) when dealing with unbalanced data. The proposed approach modifies the available training data according to the predictions of SVM and this modified training data is then used to train decision tree. As the dataset at hand i.e. COIL data is highly unbalanced with 94:6 class distribution ratios, we also employed various standard sampling techniques for extensive analysis. Based on sensitivity measure, it is observed that the proposed approach enhanced the efficiency of decision tree exceptionally well. Other intelligent methods can be tested in place of DT.