Increasing the Efficiency of Support Vector Machine by Simplifying the Shape of Separation Hypersurface

Yiqiang Zhan, Dinggang Shen

Research output: Contribution to journalArticle

Abstract

This paper presents a four-step training method for increasing the efficiency of support vector machine (SVM) by simplifying the shape of separation hypersurface. First, a SVM is initially trained by all the training samples, thereby producing a number of support vectors. Second, the support vectors, which make the hypersurface highly convoluted, are excluded from the training set. Third, the SVM is re-trained only by the remaining samples in the training set. Finally, the complexity of the trained SVM is further reduced by approximating the separation hypersurface with a subset of the support vectors. Compared to the initially trained SVM by all samples, the efficiency of the finally-trained SVM is highly improved, without system degradation.

Original languageEnglish
Pages (from-to)732-738
Number of pages7
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume3314
Publication statusPublished - 2004 Dec 1
Externally publishedYes

Fingerprint

Hypersurface
Support vector machines
Support Vector Machine
Support Vector
Training Samples
Degradation
Subset
Training

ASJC Scopus subject areas

  • Computer Science(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Theoretical Computer Science

Cite this

@article{b86248dfe42b47139c9ee012b77b4a47,
title = "Increasing the Efficiency of Support Vector Machine by Simplifying the Shape of Separation Hypersurface",
abstract = "This paper presents a four-step training method for increasing the efficiency of support vector machine (SVM) by simplifying the shape of separation hypersurface. First, a SVM is initially trained by all the training samples, thereby producing a number of support vectors. Second, the support vectors, which make the hypersurface highly convoluted, are excluded from the training set. Third, the SVM is re-trained only by the remaining samples in the training set. Finally, the complexity of the trained SVM is further reduced by approximating the separation hypersurface with a subset of the support vectors. Compared to the initially trained SVM by all samples, the efficiency of the finally-trained SVM is highly improved, without system degradation.",
author = "Yiqiang Zhan and Dinggang Shen",
year = "2004",
month = "12",
day = "1",
language = "English",
volume = "3314",
pages = "732--738",
journal = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
issn = "0302-9743",
publisher = "Springer Verlag",

}

TY - JOUR

T1 - Increasing the Efficiency of Support Vector Machine by Simplifying the Shape of Separation Hypersurface

AU - Zhan, Yiqiang

AU - Shen, Dinggang

PY - 2004/12/1

Y1 - 2004/12/1

N2 - This paper presents a four-step training method for increasing the efficiency of support vector machine (SVM) by simplifying the shape of separation hypersurface. First, a SVM is initially trained by all the training samples, thereby producing a number of support vectors. Second, the support vectors, which make the hypersurface highly convoluted, are excluded from the training set. Third, the SVM is re-trained only by the remaining samples in the training set. Finally, the complexity of the trained SVM is further reduced by approximating the separation hypersurface with a subset of the support vectors. Compared to the initially trained SVM by all samples, the efficiency of the finally-trained SVM is highly improved, without system degradation.

AB - This paper presents a four-step training method for increasing the efficiency of support vector machine (SVM) by simplifying the shape of separation hypersurface. First, a SVM is initially trained by all the training samples, thereby producing a number of support vectors. Second, the support vectors, which make the hypersurface highly convoluted, are excluded from the training set. Third, the SVM is re-trained only by the remaining samples in the training set. Finally, the complexity of the trained SVM is further reduced by approximating the separation hypersurface with a subset of the support vectors. Compared to the initially trained SVM by all samples, the efficiency of the finally-trained SVM is highly improved, without system degradation.

UR - http://www.scopus.com/inward/record.url?scp=35048825809&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35048825809&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:35048825809

VL - 3314

SP - 732

EP - 738

JO - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

JF - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SN - 0302-9743

ER -