Correcting sample selection bias by unlabeled data

Jiayuan Huang; Alexander J. Smola; Arthur Gretton; Karsten M. Borgwardt; Bernhard Schölkopf

Correcting sample selection bias by unlabeled data

Jiayuan Huang^*, Alexander J. Smola, Arthur Gretton, Karsten M. Borgwardt, Bernhard Schölkopf

^*Corresponding author for this work

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

1067 Citations (Scopus)

Abstract

We consider the scenario where training and test data are drawn from different distributions, commonly referred to as sample selection bias. Most algorithms for this setting try to first recover sampling distributions and then make appropriate corrections based on the distribution estimate. We present a nonparametric method which directly produces resampling weights without distribution estimation. Our method works by matching distributions between training and testing sets in feature space. Experimental results demonstrate that our method works well in practice.

Original language	English
Title of host publication	Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference
Pages	601-608
Number of pages	8
Publication status	Published - 2007
Event	20th Annual Conference on Neural Information Processing Systems, NIPS 2006 - Vancouver, BC, Canada Duration: 4 Dec 2006 → 7 Dec 2006

Publication series

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

Conference

Conference	20th Annual Conference on Neural Information Processing Systems, NIPS 2006
Country/Territory	Canada
City	Vancouver, BC
Period	4/12/06 → 7/12/06

Cite this

@inproceedings{62c8faedbebc4632b3de1076606f7a8c,

title = "Correcting sample selection bias by unlabeled data",

abstract = "We consider the scenario where training and test data are drawn from different distributions, commonly referred to as sample selection bias. Most algorithms for this setting try to first recover sampling distributions and then make appropriate corrections based on the distribution estimate. We present a nonparametric method which directly produces resampling weights without distribution estimation. Our method works by matching distributions between training and testing sets in feature space. Experimental results demonstrate that our method works well in practice.",

author = "Jiayuan Huang and Smola, \{Alexander J.\} and Arthur Gretton and Borgwardt, \{Karsten M.\} and Bernhard Sch{\"o}lkopf",

year = "2007",

language = "English",

isbn = "9780262195683",

series = "Advances in Neural Information Processing Systems",

pages = "601--608",

booktitle = "Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference",

note = "20th Annual Conference on Neural Information Processing Systems, NIPS 2006 ; Conference date: 04-12-2006 Through 07-12-2006",

}

Huang, J, Smola, AJ, Gretton, A, Borgwardt, KM & Schölkopf, B 2007, Correcting sample selection bias by unlabeled data. in Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference. Advances in Neural Information Processing Systems, pp. 601-608, 20th Annual Conference on Neural Information Processing Systems, NIPS 2006, Vancouver, BC, Canada, 4/12/06.

Correcting sample selection bias by unlabeled data. / Huang, Jiayuan; Smola, Alexander J.; Gretton, Arthur et al.
Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference. 2007. p. 601-608 (Advances in Neural Information Processing Systems).

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

TY - GEN

T1 - Correcting sample selection bias by unlabeled data

AU - Huang, Jiayuan

AU - Smola, Alexander J.

AU - Gretton, Arthur

AU - Borgwardt, Karsten M.

AU - Schölkopf, Bernhard

PY - 2007

Y1 - 2007

N2 - We consider the scenario where training and test data are drawn from different distributions, commonly referred to as sample selection bias. Most algorithms for this setting try to first recover sampling distributions and then make appropriate corrections based on the distribution estimate. We present a nonparametric method which directly produces resampling weights without distribution estimation. Our method works by matching distributions between training and testing sets in feature space. Experimental results demonstrate that our method works well in practice.

AB - We consider the scenario where training and test data are drawn from different distributions, commonly referred to as sample selection bias. Most algorithms for this setting try to first recover sampling distributions and then make appropriate corrections based on the distribution estimate. We present a nonparametric method which directly produces resampling weights without distribution estimation. Our method works by matching distributions between training and testing sets in feature space. Experimental results demonstrate that our method works well in practice.

UR - https://www.scopus.com/pages/publications/84864031047

M3 - Conference Paper

SN - 9780262195683

T3 - Advances in Neural Information Processing Systems

SP - 601

EP - 608

BT - Advances in Neural Information Processing Systems 19 - Proceedings of the 2006 Conference

T2 - 20th Annual Conference on Neural Information Processing Systems, NIPS 2006

Y2 - 4 December 2006 through 7 December 2006

ER -

Correcting sample selection bias by unlabeled data

Abstract

Publication series

Conference

Other files and links

Fingerprint

Cite this