ν-Arc: Ensemble learning in the presence of outliers

G. Rätsch; B. Schölkopf; A. Smola; K. R. Müller; T. Onoda; S. Mika

ν-Arc: Ensemble learning in the presence of outliers

G. Rätsch, B. Schölkopf, A. Smola, K. R. Müller, T. Onoda, S. Mika

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

Abstract

AdaBoost and other ensemble methods have successfully been applied to a number of classification tasks' seemingly defying problems of overfitting. AdaBoost performs gradient descent in an error function with respect to the margin' asymptotically concentrating on the patterns which are hardest to learn. For very noisy problems' however' this can be disadvantageous. Indeed' theoretical analysis has shown that the margin distribution' as opposed to just the minimal margin' plays a crucial role in understanding this phenomenon. Loosely speaking' some outliers should be tolerated if this has the benefit of substantially the margin on the remaining points. We propose a new boosting algorithm which allows for the possibility of a pre-specified fraction of points to lie in the margin area or even on the wrong side of the decision boundary.

Original language	English
Title of host publication	Advances in Neural Information Processing Systems 12 - Proceedings of the 1999 Conference, NIPS 1999
Publisher	Neural information processing systems foundation
Pages	561-567
Number of pages	7
ISBN (Print)	0262194503, 9780262194501
Publication status	Published - 2000
Event	13th Annual Neural Information Processing Systems Conference, NIPS 1999 - Denver, CO, United States Duration: 1999 Nov 29 → 1999 Dec 4

Publication series

Name	Advances in Neural Information Processing Systems
ISSN (Print)	1049-5258

Other

Other	13th Annual Neural Information Processing Systems Conference, NIPS 1999
Country/Territory	United States
City	Denver, CO
Period	99/11/29 → 99/12/4

ASJC Scopus subject areas

Computer Networks and Communications
Information Systems
Signal Processing

Cite this

ν-Arc: Ensemble learning in the presence of outliers. / Rätsch, G.; Schölkopf, B.; Smola, A. et al.
Advances in Neural Information Processing Systems 12 - Proceedings of the 1999 Conference, NIPS 1999. Neural information processing systems foundation, 2000. p. 561-567 (Advances in Neural Information Processing Systems).

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

Rätsch, G, Schölkopf, B, Smola, A, Müller, KR, Onoda, T & Mika, S 2000, ν-Arc: Ensemble learning in the presence of outliers. in Advances in Neural Information Processing Systems 12 - Proceedings of the 1999 Conference, NIPS 1999. Advances in Neural Information Processing Systems, Neural information processing systems foundation, pp. 561-567, 13th Annual Neural Information Processing Systems Conference, NIPS 1999, Denver, CO, United States, 99/11/29.

@inproceedings{b87f1f86d43043c18550650cbb79622e,

title = "ν-Arc: Ensemble learning in the presence of outliers",

abstract = "AdaBoost and other ensemble methods have successfully been applied to a number of classification tasks' seemingly defying problems of overfitting. AdaBoost performs gradient descent in an error function with respect to the margin' asymptotically concentrating on the patterns which are hardest to learn. For very noisy problems' however' this can be disadvantageous. Indeed' theoretical analysis has shown that the margin distribution' as opposed to just the minimal margin' plays a crucial role in understanding this phenomenon. Loosely speaking' some outliers should be tolerated if this has the benefit of substantially the margin on the remaining points. We propose a new boosting algorithm which allows for the possibility of a pre-specified fraction of points to lie in the margin area or even on the wrong side of the decision boundary.",

author = "G. R{\"a}tsch and B. Sch{\"o}lkopf and A. Smola and M{\"u}ller, {K. R.} and T. Onoda and S. Mika",

year = "2000",

language = "English",

isbn = "0262194503",

series = "Advances in Neural Information Processing Systems",

publisher = "Neural information processing systems foundation",

pages = "561--567",

booktitle = "Advances in Neural Information Processing Systems 12 - Proceedings of the 1999 Conference, NIPS 1999",

note = "13th Annual Neural Information Processing Systems Conference, NIPS 1999 ; Conference date: 29-11-1999 Through 04-12-1999",

}

TY - GEN

T1 - ν-Arc

T2 - 13th Annual Neural Information Processing Systems Conference, NIPS 1999

AU - Rätsch, G.

AU - Schölkopf, B.

AU - Smola, A.

AU - Müller, K. R.

AU - Onoda, T.

AU - Mika, S.

PY - 2000

Y1 - 2000

N2 - AdaBoost and other ensemble methods have successfully been applied to a number of classification tasks' seemingly defying problems of overfitting. AdaBoost performs gradient descent in an error function with respect to the margin' asymptotically concentrating on the patterns which are hardest to learn. For very noisy problems' however' this can be disadvantageous. Indeed' theoretical analysis has shown that the margin distribution' as opposed to just the minimal margin' plays a crucial role in understanding this phenomenon. Loosely speaking' some outliers should be tolerated if this has the benefit of substantially the margin on the remaining points. We propose a new boosting algorithm which allows for the possibility of a pre-specified fraction of points to lie in the margin area or even on the wrong side of the decision boundary.

AB - AdaBoost and other ensemble methods have successfully been applied to a number of classification tasks' seemingly defying problems of overfitting. AdaBoost performs gradient descent in an error function with respect to the margin' asymptotically concentrating on the patterns which are hardest to learn. For very noisy problems' however' this can be disadvantageous. Indeed' theoretical analysis has shown that the margin distribution' as opposed to just the minimal margin' plays a crucial role in understanding this phenomenon. Loosely speaking' some outliers should be tolerated if this has the benefit of substantially the margin on the remaining points. We propose a new boosting algorithm which allows for the possibility of a pre-specified fraction of points to lie in the margin area or even on the wrong side of the decision boundary.

UR - http://www.scopus.com/inward/record.url?scp=0008161172&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:0008161172

SN - 0262194503

SN - 9780262194501

T3 - Advances in Neural Information Processing Systems

SP - 561

EP - 567

BT - Advances in Neural Information Processing Systems 12 - Proceedings of the 1999 Conference, NIPS 1999

PB - Neural information processing systems foundation

Y2 - 29 November 1999 through 4 December 1999

ER -

ν-Arc: Ensemble learning in the presence of outliers

Abstract

Publication series

Other

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this