Research on ensemble learning based on discretization method

A novel ensemble learning algorithm based on discretization method is proposed in this paper. This algorithm uses the rough sets and Boolean reasoning approach to construct base classifiers with good diversity, which can improve the performance of ensemble learning. Then the consistency level coined from the rough sets theory is utilized to measure the information loss and control the algorithmpsilas discretization process. In combination with the support vector machine (SVM) ensemble the proposed algorithm can improve the predictive accuracy. Experimental results show that this algorithm can achieve better performance than the traditional ensemble learning methods such as Bagging and Adaboost. This algorithm can also be used in neural network ensemble.