A consistency-based model selection for one-class classification

David M J Tax, Klaus Muller

Research output: Chapter in Book/Report/Conference proceedingConference contribution

36 Citations (Scopus)

Abstract

Model selection in unsupervised learning is a hard problem. In this paper a simple selection criterion for hyperparameters in one-class classifiers (OCCs) is proposed. It makes use of the particular structure of the one-class problem. The mean idea is that the complexity of the classifier is increased until the classifier becomes inconsistent on the target class. This defines the most complex classifier which can still reliably be trained on the data. Experiments indicated the usefulness of the approach.

Original languageEnglish
Title of host publicationProceedings - International Conference on Pattern Recognition
EditorsJ. Kittler, M. Petrou, M. Nixon
Pages363-366
Number of pages4
Volume3
Publication statusPublished - 2004
Externally publishedYes
EventProceedings of the 17th International Conference on Pattern Recognition, ICPR 2004 - Cambridge, United Kingdom
Duration: 2004 Aug 232004 Aug 26

Other

OtherProceedings of the 17th International Conference on Pattern Recognition, ICPR 2004
CountryUnited Kingdom
CityCambridge
Period04/8/2304/8/26

Fingerprint

Classifiers
Unsupervised learning
Experiments

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Computer Vision and Pattern Recognition
  • Hardware and Architecture

Cite this

Tax, D. M. J., & Muller, K. (2004). A consistency-based model selection for one-class classification. In J. Kittler, M. Petrou, & M. Nixon (Eds.), Proceedings - International Conference on Pattern Recognition (Vol. 3, pp. 363-366)

A consistency-based model selection for one-class classification. / Tax, David M J; Muller, Klaus.

Proceedings - International Conference on Pattern Recognition. ed. / J. Kittler; M. Petrou; M. Nixon. Vol. 3 2004. p. 363-366.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Tax, DMJ & Muller, K 2004, A consistency-based model selection for one-class classification. in J Kittler, M Petrou & M Nixon (eds), Proceedings - International Conference on Pattern Recognition. vol. 3, pp. 363-366, Proceedings of the 17th International Conference on Pattern Recognition, ICPR 2004, Cambridge, United Kingdom, 04/8/23.
Tax DMJ, Muller K. A consistency-based model selection for one-class classification. In Kittler J, Petrou M, Nixon M, editors, Proceedings - International Conference on Pattern Recognition. Vol. 3. 2004. p. 363-366
Tax, David M J ; Muller, Klaus. / A consistency-based model selection for one-class classification. Proceedings - International Conference on Pattern Recognition. editor / J. Kittler ; M. Petrou ; M. Nixon. Vol. 3 2004. pp. 363-366
@inproceedings{6806a126d8b049d99921b0d6ba1fa52d,
title = "A consistency-based model selection for one-class classification",
abstract = "Model selection in unsupervised learning is a hard problem. In this paper a simple selection criterion for hyperparameters in one-class classifiers (OCCs) is proposed. It makes use of the particular structure of the one-class problem. The mean idea is that the complexity of the classifier is increased until the classifier becomes inconsistent on the target class. This defines the most complex classifier which can still reliably be trained on the data. Experiments indicated the usefulness of the approach.",
author = "Tax, {David M J} and Klaus Muller",
year = "2004",
language = "English",
volume = "3",
pages = "363--366",
editor = "J. Kittler and M. Petrou and M. Nixon",
booktitle = "Proceedings - International Conference on Pattern Recognition",

}

TY - GEN

T1 - A consistency-based model selection for one-class classification

AU - Tax, David M J

AU - Muller, Klaus

PY - 2004

Y1 - 2004

N2 - Model selection in unsupervised learning is a hard problem. In this paper a simple selection criterion for hyperparameters in one-class classifiers (OCCs) is proposed. It makes use of the particular structure of the one-class problem. The mean idea is that the complexity of the classifier is increased until the classifier becomes inconsistent on the target class. This defines the most complex classifier which can still reliably be trained on the data. Experiments indicated the usefulness of the approach.

AB - Model selection in unsupervised learning is a hard problem. In this paper a simple selection criterion for hyperparameters in one-class classifiers (OCCs) is proposed. It makes use of the particular structure of the one-class problem. The mean idea is that the complexity of the classifier is increased until the classifier becomes inconsistent on the target class. This defines the most complex classifier which can still reliably be trained on the data. Experiments indicated the usefulness of the approach.

UR - http://www.scopus.com/inward/record.url?scp=10044248411&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10044248411&partnerID=8YFLogxK

M3 - Conference contribution

VL - 3

SP - 363

EP - 366

BT - Proceedings - International Conference on Pattern Recognition

A2 - Kittler, J.

A2 - Petrou, M.

A2 - Nixon, M.

ER -