Ghent University Academic Bibliography

Advanced

A practical approach to model selection for support vector machines with a gaussian kernel

Matthias Varewyck UGent and Jean-Pierre Martens UGent (2011) IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS. 41(2). p.330-340
abstract
When learning a support vector machine (SVM) from a set of labeled development patterns, the ultimate goal is to get a classifier attaining a low error rate on new patterns. This so-called generalization ability obviously depends on the choices of the learning parameters that control the learning process. Model selection is the method for identifying appropriate values for these parameters. In this paper, a novel model selection method for SVMs with a Gaussian kernel is proposed. Its aim is to find suitable values for the kernel parameter. and the cost parameter C with a minimum amount of central processing unit time. The determination of the kernel parameter is based on the argument that, for most patterns, the decision function of the SVM should consist of a sufficiently large number of significant contributions. A unique property of the proposed method is that it retrieves the kernel parameter as a simple analytical function of the dimensionality of the feature space and the dispersion of the classes in that space. An experimental evaluation on a test bed of 17 classification problems has shown that the new method favorably competes with two recently published methods: the classification of new patterns is equally good, but the computational effort to identify the learning parameters is substantially lower.
Please use this url to cite or link to this publication:
author
organization
year
type
journalArticle (original)
publication status
published
subject
keyword
RECOGNITION, INITIALIZATION, Data mining, model selection, pattern classification, support vector machine (SVM)
journal title
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS
IEEE Trans. Syst. Man Cybern. Part B-Cybern.
volume
41
issue
2
pages
330 - 340
Web of Science type
Article
Web of Science id
000288454300002
JCR category
AUTOMATION & CONTROL SYSTEMS
JCR impact factor
3.08 (2011)
JCR rank
2/58 (2011)
JCR quartile
1 (2011)
ISSN
1083-4419
DOI
10.1109/TSMCB.2010.2053026
language
English
UGent publication?
yes
classification
A1
copyright statement
I have transferred the copyright for this publication to the publisher
id
1092707
handle
http://hdl.handle.net/1854/LU-1092707
date created
2010-12-23 17:42:52
date last changed
2012-05-02 14:35:57
@article{1092707,
  abstract     = {When learning a support vector machine (SVM) from a set of labeled development patterns, the ultimate goal is to get a classifier attaining a low error rate on new patterns. This so-called generalization ability obviously depends on the choices of the learning parameters that control the learning process. Model selection is the method for identifying appropriate values for these parameters. In this paper, a novel model selection method for SVMs with a Gaussian kernel is proposed. Its aim is to find suitable values for the kernel parameter. and the cost parameter C with a minimum amount of central processing unit time. The determination of the kernel parameter is based on the argument that, for most patterns, the decision function of the SVM should consist of a sufficiently large number of significant contributions. A unique property of the proposed method is that it retrieves the kernel parameter as a simple analytical function of the dimensionality of the feature space and the dispersion of the classes in that space. An experimental evaluation on a test bed of 17 classification problems has shown that the new method favorably competes with two recently published methods: the classification of new patterns is equally good, but the computational effort to identify the learning parameters is substantially lower.},
  author       = {Varewyck, Matthias and Martens, Jean-Pierre},
  issn         = {1083-4419},
  journal      = {IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS},
  keyword      = {RECOGNITION,INITIALIZATION,Data mining,model selection,pattern classification,support vector machine (SVM)},
  language     = {eng},
  number       = {2},
  pages        = {330--340},
  title        = {A practical approach to model selection for support vector machines with a gaussian kernel},
  url          = {http://dx.doi.org/10.1109/TSMCB.2010.2053026},
  volume       = {41},
  year         = {2011},
}

Chicago
Varewyck, Matthias, and Jean-Pierre Martens. 2011. “A Practical Approach to Model Selection for Support Vector Machines with a Gaussian Kernel.” Ieee Transactions on Systems Man and Cybernetics Part B-cybernetics 41 (2): 330–340.
APA
Varewyck, M., & Martens, J.-P. (2011). A practical approach to model selection for support vector machines with a gaussian kernel. IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS, 41(2), 330–340.
Vancouver
1.
Varewyck M, Martens J-P. A practical approach to model selection for support vector machines with a gaussian kernel. IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS. 2011;41(2):330–40.
MLA
Varewyck, Matthias, and Jean-Pierre Martens. “A Practical Approach to Model Selection for Support Vector Machines with a Gaussian Kernel.” IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS 41.2 (2011): 330–340. Print.