摘要
提出一种新的基于二叉树结构的支持向量(SVM)多类分类算法.该算法解决了现有主要算法所存在的不可分区域问题.为了获得较高的推广能力,必须让样本分布广的类处于二叉树的上层节点,才能获得更大的划分空间.所以,该算法采用最小超立方体和最小超球体类包含作为二叉树的生成算法.实验结果表明,该算法具有一定的优越性.
The multiclass SVM methods based on binary tree are proposed. The new methods can resolve the unclassifiable region problems in the conventional multiclass SVM methods. To maintain high generalization ability, the most widespread class should be separated at the upper nodes of a binary tree. Hypercuboid and hypersphere class least covers are used to be rules of constructing binary tree. Numerical experiment results show that the multiclass SVM methods are suitable for practical use.
出处
《控制与决策》
EI
CSCD
北大核心
2005年第7期746-749,754,共5页
Control and Decision
基金
国家自然科学基金项目(79970025)
国防科技预研基金项目(00J15.3.3.JW0528).
关键词
支持向量机
多类分类
二叉树
多类支持向量机
Support vector machines
Multiclass classification
Binary tree
Multiclass support vector machines
