一种基于双支持向量机决策树的多分类算法

he twin support vector machine (TSVM) can effectively reduce the training time compared with the traditional support vector machine (SVM).While the SVM decision tree has a good performance in solving the multi-class classification .In this paper, the TSVM and the decision tree method are combined, and we use the dissimilarity between classes to construct an optimized TSVM decision tree (DT-TSVM). Finally the experiment based on the UCI data sets prove that our DT-TSVM not only has a higher classification accuracy rate but also has shorter training time.