A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data

Xiaoyu Ding; Jong Hwan Lee; Seong Whan Lee

doi:10.1109/ICMLC.2011.6016680

A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data

Xiaoyu Ding, Jong Hwan Lee, Seong Whan Lee

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

6 Citations (Scopus)

Abstract

This paper proposes a constrained alternating least squares nonnegative matrix factorization algorithm (cALSNMF) to enhance alternating least squares non-negative matrix factorization (ALSNMF) in detecting task-related neuronal activity from single subject's fMRI data. In cALSNMF, a new cost function is defined in consideration of the uncorrelation and overdeter-mined problems of fMRI data, A novel training procedure is generated by combining optimal brain surgeon (OBS) algorithm in weight updating process, which considers the interaction among voxels. The experiments on both simulated data and fMRI data show that cALSNMF fits data better without any prior information and works more adaptively than original ALSNMF on detecting task-related neuronal activity.

Original language	English
Title of host publication	Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011
Pages	338-343
Number of pages	6
DOIs	https://doi.org/10.1109/ICMLC.2011.6016680
Publication status	Published - 2011
Event	2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011 - Guilin, Guangxi, China Duration: 2011 Jul 10 → 2011 Jul 13

Publication series

Name	Proceedings - International Conference on Machine Learning and Cybernetics
Volume	1
ISSN (Print)	2160-133X
ISSN (Electronic)	2160-1348

Other

Other	2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011
Country/Territory	China
City	Guilin, Guangxi
Period	11/7/10 → 11/7/13

Keywords

Constrained alternating least squares nonnegative matrix factorization
fMRI
optimal brain surgeon

ASJC Scopus subject areas

Artificial Intelligence
Computational Theory and Mathematics
Computer Networks and Communications
Human-Computer Interaction

Access to Document

10.1109/ICMLC.2011.6016680

Cite this

Ding, X., Lee, J. H., & Lee, S. W. (2011). A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data. In Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011 (pp. 338-343). Article 6016680 (Proceedings - International Conference on Machine Learning and Cybernetics; Vol. 1). https://doi.org/10.1109/ICMLC.2011.6016680

A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data. / Ding, Xiaoyu; Lee, Jong Hwan ; Lee, Seong Whan.
Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011. 2011. p. 338-343 6016680 (Proceedings - International Conference on Machine Learning and Cybernetics; Vol. 1).

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

Ding, X, Lee, JH & Lee, SW 2011, A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data. in Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011., 6016680, Proceedings - International Conference on Machine Learning and Cybernetics, vol. 1, pp. 338-343, 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011, Guilin, Guangxi, China, 11/7/10. https://doi.org/10.1109/ICMLC.2011.6016680

Ding X, Lee JH , Lee SW. A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data. In Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011. 2011. p. 338-343. 6016680. (Proceedings - International Conference on Machine Learning and Cybernetics). doi: 10.1109/ICMLC.2011.6016680

Ding, Xiaoyu ; Lee, Jong Hwan ; Lee, Seong Whan. / A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data. Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011. 2011. pp. 338-343 (Proceedings - International Conference on Machine Learning and Cybernetics).

@inproceedings{a1c3db3deab74de696747b4e34cc4295,

title = "A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data",

abstract = "This paper proposes a constrained alternating least squares nonnegative matrix factorization algorithm (cALSNMF) to enhance alternating least squares non-negative matrix factorization (ALSNMF) in detecting task-related neuronal activity from single subject's fMRI data. In cALSNMF, a new cost function is defined in consideration of the uncorrelation and overdeter-mined problems of fMRI data, A novel training procedure is generated by combining optimal brain surgeon (OBS) algorithm in weight updating process, which considers the interaction among voxels. The experiments on both simulated data and fMRI data show that cALSNMF fits data better without any prior information and works more adaptively than original ALSNMF on detecting task-related neuronal activity.",

keywords = "Constrained alternating least squares nonnegative matrix factorization, fMRI, optimal brain surgeon",

author = "Xiaoyu Ding and Lee, {Jong Hwan} and Lee, {Seong Whan}",

year = "2011",

doi = "10.1109/ICMLC.2011.6016680",

language = "English",

isbn = "9781457703065",

series = "Proceedings - International Conference on Machine Learning and Cybernetics",

pages = "338--343",

booktitle = "Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011",

note = "2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011 ; Conference date: 10-07-2011 Through 13-07-2011",

}

TY - GEN

T1 - A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data

AU - Ding, Xiaoyu

AU - Lee, Jong Hwan

AU - Lee, Seong Whan

PY - 2011

Y1 - 2011

N2 - This paper proposes a constrained alternating least squares nonnegative matrix factorization algorithm (cALSNMF) to enhance alternating least squares non-negative matrix factorization (ALSNMF) in detecting task-related neuronal activity from single subject's fMRI data. In cALSNMF, a new cost function is defined in consideration of the uncorrelation and overdeter-mined problems of fMRI data, A novel training procedure is generated by combining optimal brain surgeon (OBS) algorithm in weight updating process, which considers the interaction among voxels. The experiments on both simulated data and fMRI data show that cALSNMF fits data better without any prior information and works more adaptively than original ALSNMF on detecting task-related neuronal activity.

AB - This paper proposes a constrained alternating least squares nonnegative matrix factorization algorithm (cALSNMF) to enhance alternating least squares non-negative matrix factorization (ALSNMF) in detecting task-related neuronal activity from single subject's fMRI data. In cALSNMF, a new cost function is defined in consideration of the uncorrelation and overdeter-mined problems of fMRI data, A novel training procedure is generated by combining optimal brain surgeon (OBS) algorithm in weight updating process, which considers the interaction among voxels. The experiments on both simulated data and fMRI data show that cALSNMF fits data better without any prior information and works more adaptively than original ALSNMF on detecting task-related neuronal activity.

KW - Constrained alternating least squares nonnegative matrix factorization

KW - fMRI

KW - optimal brain surgeon

UR - http://www.scopus.com/inward/record.url?scp=80155143321&partnerID=8YFLogxK

U2 - 10.1109/ICMLC.2011.6016680

DO - 10.1109/ICMLC.2011.6016680

M3 - Conference contribution

AN - SCOPUS:80155143321

SN - 9781457703065

T3 - Proceedings - International Conference on Machine Learning and Cybernetics

SP - 338

EP - 343

BT - Proceedings of 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011

T2 - 2011 International Conference on Machine Learning and Cybernetics, ICMLC 2011

Y2 - 10 July 2011 through 13 July 2011

ER -

A constrained alternating least squares nonnegative matrix factorization algorithm enhances task-related neuronal activity detection from single subject's fMRI data

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this