基于粗糙集边界的v-支持向量机混合分类算法

针对v-支持向量机在样本集规模较大的情况下,需要占用大量训练时间的问题,提出基于粗糙集边界的v-支持向量机混合分类算法。该算法根据粗糙集理论边界区域的优点,生成分类数据的边界集,使其包括全部的支持向量,用此边界向量集替代原始样本作为训练集,减少训练集的数量,则可以在不影响分类精度和泛化性能的前提下显著缩短v-支持向量机的训练时间。仿真结果表明该算法的有效性。

V-Support Vector Machine Hybrid Classification Algorithm Based on Boundary of Rough Set

V-support vector machine（v-SVM） can take up a lot of training time when large-scale samples set.V-support vector machine hybrid classification algorithm based on boundary of rough set（RSBv-SVM） is proposed.According to the merits of boundary region of rough set theory,the algorithm gets the boundary set of the classified data,which includes all support vectors.The boundary set can substitute the original inputs as a training subset,and the size of the training set is decreased.Training time is reduced by v-SVM while keeping the accuracy of classification and the performance of generalization.The simulation experiments show the effectiveness of the suggested hybrid method.