论文部分内容阅读
为了解决多类支持向量机的选型问题,降低多类分类问题的难度,对4种常用的多类支持向量机进行了对比研究。从多类支持向量机的构造原理出发,对多类支持向量机的训练复杂度、测试复杂度和分类准确率进行了理论分析。在此基础上,利用标准数据集对多类支持向量机进行试验分析,结果表明,导向无环图支持向量机的分类准确率最高,二叉树支持向量机的实时性最优。
In order to solve the problem of multi-class SVM selection and reduce the difficulty of multi-class classification problems, four commonly used multi-class SVMs are compared. Based on the construction principle of multiple SVMs, the training complexity, test complexity and classification accuracy of multiple SVMs are theoretically analyzed. Based on this, we use the standard dataset to test the multi-class support vector machines. The results show that the classification accuracy of the guided acyclic graph support vector machine is the highest, and the binary tree support vector machine has the best real-time performance.