Application of Hadamard ECOC in Multi-Class Problems Based on SVM

YIN An-rong; XIE Xiang; KUANG Jing-ming

您当前的位置：

首页 >

文章列表页 >

Application of Hadamard ECOC in Multi-Class Problems Based on SVM

更新时间：2025-07-16

- Application of Hadamard ECOC in Multi-Class Problems Based on SVM
- Acta Electronica Sinica Vol. 36, Issue 1, Pages: 122-126(2008)
- 作者机构：
  
  北京理工大学电子工程系,北京,100081
- 作者简介：
- 基金信息：
- DOI：
  CLC： TN912.34
- Published：2008
- 稿件说明：
移动端阅览
YIN An-rong, XIE Xiang, KUANG Jing-ming. Application of Hadamard ECOC in Multi-Class Problems Based on SVM[J]. Acta Electronica Sinica, 2008, 36(1): 122-126.
DOI：

YIN An-rong, XIE Xiang, KUANG Jing-ming. Application of Hadamard ECOC in Multi-Class Problems Based on SVM[J]. Acta Electronica Sinica, 2008, 36(1): 122-126. DOI：

摘要

多分类问题一直是模式识别领域的一个热点

本文提出了将Hadamard纠错码同二元分类器相结合的方法来解决此问题

相对于其它类型的纠错码多分类器法

该方法的实现简单快捷

且更容易构造出性能优越的纠错码本.本文将Hadamard纠错码和支持向量机相结合

应用于说话人辨认这样一个多分类问题中

并同传统的"1对余"的多类推广方式进行了比较.实验结果表明在多分类任务中

Hadamard纠错码对于不同的类别都表现出了很强的分类能力

且性能优于"1对余"法

对于类间码字的不同分配方式也具有良好的鲁棒性.

Abstract

In this paper

we proposed applying Hadamard Error-Correcting Output Code to extend binary classifier to multi-class problems.Compared with other ECOC approaches

Hadamard ECOC is easy to construct and suitable to any number of classes.We combine it with binary support vector machine (SVM) to solve the multi-class problem of speaker identification.Compared with the traditional "1-against-rest" method

the experiment result shows that Hadamard ECOC has much better and more stable performance to any number of classes for the multi-class problem and is robust with respect to the assignment of distributed representations to particular classes.

关键词

Keywords

references

Views

1216

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

High Correct Recognition Rate Classifier Design with Appropriate Rejection Mechanism

A Simplification to Support Vector Machine for Complicated Recognition Problem

Some New Developments on Support Vector Machine

Structure-Wise Feature Reconstruction for Hyperspectral Image Classification

Related Author

YANG Guo-wei

WAN Ming-hua

LAI Zhi-hui

ZHANG Fan-long

WONG Wai-keung

FANG Jing-long

CHEN Shuo

PAN Zhi-geng

Related Institution

School of Electronic Information, Qingdao University

School of Information Engineering, Nanjing Audit University

School of Computer Science and Software Engineering, Shenzhen University

Hong Kong Polytechnic University, Kowloon

Institute of Graphics and Image,Hangzhou Dianzi University

⁰