Cross-domain sparse coding

Jim Jing Yan Wang, Halima Bensmail

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

2 Citations (Scopus)

Abstract

Sparse coding has shown its power as an effective data representation method. However, up to now, all the sparse coding approaches are limited within the single domain learning problem. In this paper, we extend the sparse coding to cross domain learning problem, which tries to learn from a source domain to a target domain with significant different distribution. We impose the Maximum Mean Discrepancy (MMD) criterion to reduce the cross-domain distribution difference of sparse codes, and also regularize the sparse codes by the class labels of the samples from both domains to increase the discriminative ability. The encouraging experiment results of the proposed cross-domain sparse coding algorithm on two challenging tasks - image classification of photograph and oil painting domains, and multiple user spam detection - show the advantage of the proposed method over other cross-domain data representation methods.

Original languageEnglish
Title of host publicationInternational Conference on Information and Knowledge Management, Proceedings
Pages1461-1464
Number of pages4
DOIs
Publication statusPublished - 11 Dec 2013
Event22nd ACM International Conference on Information and Knowledge Management, CIKM 2013 - San Francisco, CA, United States
Duration: 27 Oct 20131 Nov 2013

Other

Other22nd ACM International Conference on Information and Knowledge Management, CIKM 2013
CountryUnited States
CitySan Francisco, CA
Period27/10/131/11/13

Fingerprint

Spam
Oil
Discrepancy
Experiment

Keywords

  • Cross-domain learning
  • Maximum Mean Discrepancy
  • Sparse coding

ASJC Scopus subject areas

  • Business, Management and Accounting(all)
  • Decision Sciences(all)

Cite this

Wang, J. J. Y., & Bensmail, H. (2013). Cross-domain sparse coding. In International Conference on Information and Knowledge Management, Proceedings (pp. 1461-1464) https://doi.org/10.1145/2505515.2507819

Cross-domain sparse coding. / Wang, Jim Jing Yan; Bensmail, Halima.

International Conference on Information and Knowledge Management, Proceedings. 2013. p. 1461-1464.

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

Wang, JJY & Bensmail, H 2013, Cross-domain sparse coding. in International Conference on Information and Knowledge Management, Proceedings. pp. 1461-1464, 22nd ACM International Conference on Information and Knowledge Management, CIKM 2013, San Francisco, CA, United States, 27/10/13. https://doi.org/10.1145/2505515.2507819
Wang JJY, Bensmail H. Cross-domain sparse coding. In International Conference on Information and Knowledge Management, Proceedings. 2013. p. 1461-1464 https://doi.org/10.1145/2505515.2507819
Wang, Jim Jing Yan ; Bensmail, Halima. / Cross-domain sparse coding. International Conference on Information and Knowledge Management, Proceedings. 2013. pp. 1461-1464
@inproceedings{00b9e94903f74e16813e60f641c48e51,
title = "Cross-domain sparse coding",
abstract = "Sparse coding has shown its power as an effective data representation method. However, up to now, all the sparse coding approaches are limited within the single domain learning problem. In this paper, we extend the sparse coding to cross domain learning problem, which tries to learn from a source domain to a target domain with significant different distribution. We impose the Maximum Mean Discrepancy (MMD) criterion to reduce the cross-domain distribution difference of sparse codes, and also regularize the sparse codes by the class labels of the samples from both domains to increase the discriminative ability. The encouraging experiment results of the proposed cross-domain sparse coding algorithm on two challenging tasks - image classification of photograph and oil painting domains, and multiple user spam detection - show the advantage of the proposed method over other cross-domain data representation methods.",
keywords = "Cross-domain learning, Maximum Mean Discrepancy, Sparse coding",
author = "Wang, {Jim Jing Yan} and Halima Bensmail",
year = "2013",
month = "12",
day = "11",
doi = "10.1145/2505515.2507819",
language = "English",
isbn = "9781450322638",
pages = "1461--1464",
booktitle = "International Conference on Information and Knowledge Management, Proceedings",

}

TY - GEN

T1 - Cross-domain sparse coding

AU - Wang, Jim Jing Yan

AU - Bensmail, Halima

PY - 2013/12/11

Y1 - 2013/12/11

N2 - Sparse coding has shown its power as an effective data representation method. However, up to now, all the sparse coding approaches are limited within the single domain learning problem. In this paper, we extend the sparse coding to cross domain learning problem, which tries to learn from a source domain to a target domain with significant different distribution. We impose the Maximum Mean Discrepancy (MMD) criterion to reduce the cross-domain distribution difference of sparse codes, and also regularize the sparse codes by the class labels of the samples from both domains to increase the discriminative ability. The encouraging experiment results of the proposed cross-domain sparse coding algorithm on two challenging tasks - image classification of photograph and oil painting domains, and multiple user spam detection - show the advantage of the proposed method over other cross-domain data representation methods.

AB - Sparse coding has shown its power as an effective data representation method. However, up to now, all the sparse coding approaches are limited within the single domain learning problem. In this paper, we extend the sparse coding to cross domain learning problem, which tries to learn from a source domain to a target domain with significant different distribution. We impose the Maximum Mean Discrepancy (MMD) criterion to reduce the cross-domain distribution difference of sparse codes, and also regularize the sparse codes by the class labels of the samples from both domains to increase the discriminative ability. The encouraging experiment results of the proposed cross-domain sparse coding algorithm on two challenging tasks - image classification of photograph and oil painting domains, and multiple user spam detection - show the advantage of the proposed method over other cross-domain data representation methods.

KW - Cross-domain learning

KW - Maximum Mean Discrepancy

KW - Sparse coding

UR - http://www.scopus.com/inward/record.url?scp=84889603884&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84889603884&partnerID=8YFLogxK

U2 - 10.1145/2505515.2507819

DO - 10.1145/2505515.2507819

M3 - Conference contribution

SN - 9781450322638

SP - 1461

EP - 1464

BT - International Conference on Information and Knowledge Management, Proceedings

ER -