首页 | 本学科首页   官方微博 | 高级检索  
     

基于Huffman树的多类SVM方法
引用本文:刘洋,张秋余. 基于Huffman树的多类SVM方法[J]. 计算机工程与设计, 2008, 29(7): 1792-1794
作者姓名:刘洋  张秋余
作者单位:渤海大学信息科学与工程学院,辽宁锦州,121000;兰州理工大学计算机与通信学院,甘肃兰州,730050;兰州理工大学计算机与通信学院,甘肃兰州,730050
摘    要:提出了一种基于Huffman树的支持向量机多类分类方法.二叉树方法是一种常用的多类分类方法,它的关键问题在于如何构造合理的结构以获得较高的推广能力.为解决该问题,按照Huffman树的构造过程自下向上地构造二又树,使易于分割的类处于上层结点.实验结果表明,该方法与One-vs-One和DAGSVM方法的分类效果相当.

关 键 词:机器学习  支持向量机  多类分类  二叉树  哈夫曼树
文章编号:1000-7024(2008)07-1792-02
修稿时间:2007-10-08

Multiclass SVM method based on Huffman tree
LIU Yang,ZHANG Qiu-yu. Multiclass SVM method based on Huffman tree[J]. Computer Engineering and Design, 2008, 29(7): 1792-1794
Authors:LIU Yang  ZHANG Qiu-yu
Affiliation:LIU Yang1,2,ZHANG Qiu-yu2(1.College of Information Science , Engineering,Bohai University,Jinzhou 121000,China,2.College of Computer , Communication,Lanzhou University of Technology,Lanzhou 730050,China)
Abstract:A support vector machine multiclass classification method based on Huffman tree is put forward.The binary tree method of multiclass classification is an ordinary method,the key issue of which is how to construct a reasonable structure to maintain high gene-ralization ability.To solve the problem,a binary tree is constructed from down to up according to the construction process of Huffman tree,making the class easy to separate lies in the upper node.Numerical experiments show that the method proposed is comp...
Keywords:machine learning  support vector machine  multiclass classification  binary tree  Huffman tree  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号