Incremental support vector machine learning: A local approach

Liva Ralaivola; Florence D’alché-Buc

doi:10.1007/3-540-44668-0_46

Incremental support vector machine learning: A local approach

Liva Ralaivola
, Florence D’alché-Buc

LIP6, UPMC Sorbonne Universités - Paris 6

Résultats de recherche: Le chapitre dans un livre, un rapport, une anthologie ou une collection › Contribution à une conférence › Revue par des pairs

Résumé

In this paper, we propose and study a new on-line algorithm for learning a SVM based on Radial Basis Function Kernel: Local Incremental Learning of SVM or LISVM. Our method exploits the “locality” of RBF kernels to update current machine by only considering a subset of support candidates in the neighbourhood of the input. The determination of this subset is conditioned by the computation of the variation of the error estimate. Implementation is based on the SMO one, introduced and developed by Platt [13]. We study the behaviour of the algorithm during learning when using different generalization error estimates. Experiments on three data sets (batch problems transformed into on-line ones) have been conducted and analyzed.

langue originale	Anglais
titre	Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings
rédacteurs en chef	Kurt Hornik, Georg Dorffner, Horst Bischof
Editeur	Springer Verlag
Pages	322-330
Nombre de pages	9
ISBN (imprimé)	3540424865, 9783540446682
Les DOIs	https://doi.org/10.1007/3-540-44668-0_46
état	Publié - 1 janv. 2001
Modification externe	Oui
Evénement	International Conference on Artificial Neural Networks, ICANN 2001 - Vienna, Autriche Durée: 21 août 2001 → 25 août 2001

Série de publications

Nom	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	2130
ISSN (imprimé)	0302-9743
ISSN (Electronique)	1611-3349

Une conférence

Une conférence	International Conference on Artificial Neural Networks, ICANN 2001
Pays/Territoire	Autriche
La ville	Vienna
période	21/08/01 → 25/08/01

Accès au document

10.1007/3-540-44668-0_46

Contient cette citation

Ralaivola, L., & D’alché-Buc, F. (2001). Incremental support vector machine learning: A local approach. Dans K. Hornik, G. Dorffner, & H. Bischof (eds.), Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings (p. 322-330). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol 2130). Springer Verlag. https://doi.org/10.1007/3-540-44668-0_46

Ralaivola, Liva ; D’alché-Buc, Florence. / Incremental support vector machine learning : A local approach. Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. Editeur / Kurt Hornik ; Georg Dorffner ; Horst Bischof. Springer Verlag, 2001. p. 322-330 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{975bdf314a5449ebad9bcd3d357910b0,

title = "Incremental support vector machine learning: A local approach",

abstract = "In this paper, we propose and study a new on-line algorithm for learning a SVM based on Radial Basis Function Kernel: Local Incremental Learning of SVM or LISVM. Our method exploits the “locality” of RBF kernels to update current machine by only considering a subset of support candidates in the neighbourhood of the input. The determination of this subset is conditioned by the computation of the variation of the error estimate. Implementation is based on the SMO one, introduced and developed by Platt [13]. We study the behaviour of the algorithm during learning when using different generalization error estimates. Experiments on three data sets (batch problems transformed into on-line ones) have been conducted and analyzed.",

author = "Liva Ralaivola and Florence D{\textquoteright}alch{\'e}-Buc",

note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 2001.; International Conference on Artificial Neural Networks, ICANN 2001 ; Conference date: 21-08-2001 Through 25-08-2001",

year = "2001",

month = jan,

day = "1",

doi = "10.1007/3-540-44668-0\_46",

language = "English",

isbn = "3540424865",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "322--330",

editor = "Kurt Hornik and Georg Dorffner and Horst Bischof",

booktitle = "Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings",

}

Ralaivola, L & D’alché-Buc, F 2001, Incremental support vector machine learning: A local approach. Dans K Hornik, G Dorffner & H Bischof (eds), Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), VOL. 2130, Springer Verlag, p. 322-330, International Conference on Artificial Neural Networks, ICANN 2001, Vienna, Autriche, 21/08/01. https://doi.org/10.1007/3-540-44668-0_46

Incremental support vector machine learning: A local approach. / Ralaivola, Liva; D’alché-Buc, Florence.
Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. Ed. / Kurt Hornik; Georg Dorffner; Horst Bischof. Springer Verlag, 2001. p. 322-330 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol 2130).

Résultats de recherche: Le chapitre dans un livre, un rapport, une anthologie ou une collection › Contribution à une conférence › Revue par des pairs

TY - GEN

T1 - Incremental support vector machine learning

T2 - International Conference on Artificial Neural Networks, ICANN 2001

AU - Ralaivola, Liva

AU - D’alché-Buc, Florence

N1 - Publisher Copyright: © Springer-Verlag Berlin Heidelberg 2001.

PY - 2001/1/1

Y1 - 2001/1/1

N2 - In this paper, we propose and study a new on-line algorithm for learning a SVM based on Radial Basis Function Kernel: Local Incremental Learning of SVM or LISVM. Our method exploits the “locality” of RBF kernels to update current machine by only considering a subset of support candidates in the neighbourhood of the input. The determination of this subset is conditioned by the computation of the variation of the error estimate. Implementation is based on the SMO one, introduced and developed by Platt [13]. We study the behaviour of the algorithm during learning when using different generalization error estimates. Experiments on three data sets (batch problems transformed into on-line ones) have been conducted and analyzed.

AB - In this paper, we propose and study a new on-line algorithm for learning a SVM based on Radial Basis Function Kernel: Local Incremental Learning of SVM or LISVM. Our method exploits the “locality” of RBF kernels to update current machine by only considering a subset of support candidates in the neighbourhood of the input. The determination of this subset is conditioned by the computation of the variation of the error estimate. Implementation is based on the SMO one, introduced and developed by Platt [13]. We study the behaviour of the algorithm during learning when using different generalization error estimates. Experiments on three data sets (batch problems transformed into on-line ones) have been conducted and analyzed.

U2 - 10.1007/3-540-44668-0_46

DO - 10.1007/3-540-44668-0_46

M3 - Conference contribution

AN - SCOPUS:84958962423

SN - 3540424865

SN - 9783540446682

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 322

EP - 330

BT - Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings

A2 - Hornik, Kurt

A2 - Dorffner, Georg

A2 - Bischof, Horst

PB - Springer Verlag

Y2 - 21 August 2001 through 25 August 2001

ER -

Ralaivola L, D’alché-Buc F. Incremental support vector machine learning: A local approach. Dans Hornik K, Dorffner G, Bischof H, rédacteurs en chef, Artificial Neural Networks - ICANN 2001 - International Conference, Proceedings. Springer Verlag. 2001. p. 322-330. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/3-540-44668-0_46

Incremental support vector machine learning: A local approach

Résumé

Série de publications

Une conférence

Accès au document

Empreinte digitale

Contient cette citation