摘要
针对传统的基于支持向量机的文本分类存在的问题进行了研究,采用二叉树结构与多个支持向量机子分类器组合进行Web文本信息分类,在二叉树支持向量机多类方法的基础上,进一步结合遗传算法,提出了一种新的支持向量机的多类分类方法,实验结果表明,采用该方法进行多类分类,分类精度明显提高,体现了将遗传算法与二叉树支持向量机结合的优越性。
We discuss the traditional support vector machine-based text classification problems,the method constructing and combining several binary support vector machines with a binary tree can solve web text classification problems,applying binary tree into support vector machine,further combination of genetic algorithm,proposes a new support vector machine multi-class classification methods, Experiment results show that the method of multi-class classification,classification accuracy improved,manifests the algorithm＇s advantage of genetic algorithm integrated with binary tree support vector machine.
出处
《长春理工大学学报（自然科学版）》
2010年第1期152-155,共4页
Journal of Changchun University of Science and Technology(Natural Science Edition)
基金
黑龙江省研究生创新科研资金项目(YJSCX2006-38HLJ)
关键词
遗传算法
多分类
支持向量机
二叉树
genetic algorithm
multi-class classification
support vector machine
binary tree
