多类支持向量机算法综述  

Multi-class Support Vector Machines algorithm summarization

作　　者：黄勇[1] 郑春颖[1] 宋忠虎[1] 

机构地区：[1]空军工程大学导弹学院,陕西三原713800

出　　处：《计算技术与自动化》2005年第4期61-63,共3页Computing Technology and Automation

基　　金：陕西省自然科学研究项目(2004F36)

摘　　要：传统的支持向量机是基于两类问题提出的,如何将其有效的推广至多类问题仍是一个有待研究的问题。本文中作者致力于对现有的几种较有成效的多类支持向量机做一介绍,并比较其优劣,以期对研究者以后的研究能有所启发。Traditional Support Vector Machines（SVM） were originally designed for binary classification. How to effectively extend it for multi - class classification is still an on - going research issue. This paper will typically introduce several existing multi - class classifier and compare them advantage and disadvantage. The author hopes that this article can give investigators some illumination in their present - day investigation.

关 键 词：支持向量机 多类 有向无环图 纠错编码支持向量机 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]