Selecting from an infinite set of features in SVM

Rémi Flamary; Florian Yger; Alain Rakotomamonjy

Selecting from an infinite set of features in SVM

Rémi Flamary
, Florian Yger
, Alain Rakotomamonjy

Normandie Université

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

4 Citations (Scopus)

Abstract

Dealing with the continuous parameters of a feature extraction method has always been a difficult task that is usually solved by cross-validation. In this paper, we propose an active set algorithm for selecting automatically these parameters in a SVM classification context. Our experiments on texture recognition and BCI signal classification show that optimizing the feature parameters in a continuous space while learning the decision function yields to better performances than using fixed parameters obtained from a grid sampling.

Original language	English
Title of host publication	Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010
Pages	327-332
Number of pages	6
Publication status	Published - 1 Dec 2010
Externally published	Yes

Publication series

Name	ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning

Cite this

Flamary, R., Yger, F., & Rakotomamonjy, A. (2010). Selecting from an infinite set of features in SVM. In Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010 (pp. 327-332). (ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning).

Flamary, Rémi ; Yger, Florian ; Rakotomamonjy, Alain. / Selecting from an infinite set of features in SVM. Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010. 2010. pp. 327-332 (ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning).

@inproceedings{f3d47a6176044a76b4f9a0e0739d1681,

title = "Selecting from an infinite set of features in SVM",

abstract = "Dealing with the continuous parameters of a feature extraction method has always been a difficult task that is usually solved by cross-validation. In this paper, we propose an active set algorithm for selecting automatically these parameters in a SVM classification context. Our experiments on texture recognition and BCI signal classification show that optimizing the feature parameters in a continuous space while learning the decision function yields to better performances than using fixed parameters obtained from a grid sampling.",

author = "R{\'e}mi Flamary and Florian Yger and Alain Rakotomamonjy",

year = "2010",

month = dec,

day = "1",

language = "English",

isbn = "9782874190445",

series = "ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning",

pages = "327--332",

booktitle = "Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010",

}

Flamary, R, Yger, F & Rakotomamonjy, A 2010, Selecting from an infinite set of features in SVM. in Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010. ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning, pp. 327-332.

Selecting from an infinite set of features in SVM. / Flamary, Rémi; Yger, Florian; Rakotomamonjy, Alain.
Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010. 2010. p. 327-332 (ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Selecting from an infinite set of features in SVM

AU - Flamary, Rémi

AU - Yger, Florian

AU - Rakotomamonjy, Alain

PY - 2010/12/1

Y1 - 2010/12/1

N2 - Dealing with the continuous parameters of a feature extraction method has always been a difficult task that is usually solved by cross-validation. In this paper, we propose an active set algorithm for selecting automatically these parameters in a SVM classification context. Our experiments on texture recognition and BCI signal classification show that optimizing the feature parameters in a continuous space while learning the decision function yields to better performances than using fixed parameters obtained from a grid sampling.

AB - Dealing with the continuous parameters of a feature extraction method has always been a difficult task that is usually solved by cross-validation. In this paper, we propose an active set algorithm for selecting automatically these parameters in a SVM classification context. Our experiments on texture recognition and BCI signal classification show that optimizing the feature parameters in a continuous space while learning the decision function yields to better performances than using fixed parameters obtained from a grid sampling.

M3 - Conference contribution

AN - SCOPUS:84887046505

SN - 9782874190445

T3 - ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning

SP - 327

EP - 332

BT - Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010

ER -

Flamary R, Yger F, Rakotomamonjy A. Selecting from an infinite set of features in SVM. In Proceedings of the 18th European Symposium on Artificial Neural Networks - Computational Intelligence and Machine Learning, ESANN 2010. 2010. p. 327-332. (ESANN 2011 proceedings, 19th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning).

Selecting from an infinite set of features in SVM

Abstract

Publication series

Fingerprint

Cite this