Transferable Feature Selection for Unsupervised Domain Adaptation: Extended Abstract

Yuguang Yan; Hanrui Wu; Yuzhong Ye; Chaoyang Bi; Min Lu; Dapeng Liu; Qingyao Wu; Michael K. Ng

doi:10.1109/ICDE55515.2023.00354

Transferable Feature Selection for Unsupervised Domain Adaptation: Extended Abstract

Yuguang Yan, Hanrui Wu, Yuzhong Ye, Chaoyang Bi, Min Lu, Dapeng Liu, Qingyao Wu, Michael K. Ng

Department of Mathematics

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

1 Citation (Scopus)

Abstract

Domain adaptation aims at extracting knowledge from auxiliary source domains to assist the learning task in a target domain. Since the distributions of the source and target domains are different, directly using source data to build a classifier for the target domain may hamper the classification performance on the target data. In this paper, we propose to find a feature subset that is both transferable and discriminative, so that both the domain discrepancy and the classification loss measured on the selected features can be reduced. To achieve this, we formulate a new sparse learning model that is able to jointly reduce the domain discrepancy and select informative features for classification. Extensive experiments on real-world data sets demonstrate the effectiveness of the proposed method.

Original language	English
Title of host publication	Proceedings - 2023 IEEE 39th International Conference on Data Engineering, ICDE 2023
Publisher	IEEE
Pages	3855-3856
Number of pages	2
ISBN (Electronic)	9798350322279
ISBN (Print)	9798350322286
DOIs	https://doi.org/10.1109/ICDE55515.2023.00354
Publication status	Published - 3 Apr 2023
Event	39th IEEE International Conference on Data Engineering, ICDE 2023 - Anaheim, United States Duration: 3 Apr 2023 → 7 Apr 2023 https://icde2023.ics.uci.edu/ https://ieeexplore.ieee.org/xpl/conhome/10184508/proceeding

Publication series

Name	Proceedings - International Conference on Data Engineering
Volume	2023-April
ISSN (Print)	1063-6382
ISSN (Electronic)	2375-026X

Competition

Competition	39th IEEE International Conference on Data Engineering, ICDE 2023
Country/Territory	United States
City	Anaheim
Period	3/04/23 → 7/04/23
Internet address	https://icde2023.ics.uci.edu/ https://ieeexplore.ieee.org/xpl/conhome/10184508/proceeding

Scopus Subject Areas

Software
Signal Processing
Information Systems

User-Defined Keywords

Domain Adaptation
Feature Selection
Sparse Learning Model

Access to Document

10.1109/ICDE55515.2023.00354

Cite this

Yan, Y., Wu, H., Ye, Y., Bi, C., Lu, M., Liu, D., Wu, Q., & Ng, M. K. (2023). Transferable Feature Selection for Unsupervised Domain Adaptation: Extended Abstract. In Proceedings - 2023 IEEE 39th International Conference on Data Engineering, ICDE 2023 (pp. 3855-3856). (Proceedings - International Conference on Data Engineering; Vol. 2023-April). IEEE. https://doi.org/10.1109/ICDE55515.2023.00354

@inproceedings{ce5fcf864a7043cbb0495b0debcdbc54,

title = "Transferable Feature Selection for Unsupervised Domain Adaptation: Extended Abstract",

abstract = "Domain adaptation aims at extracting knowledge from auxiliary source domains to assist the learning task in a target domain. Since the distributions of the source and target domains are different, directly using source data to build a classifier for the target domain may hamper the classification performance on the target data. In this paper, we propose to find a feature subset that is both transferable and discriminative, so that both the domain discrepancy and the classification loss measured on the selected features can be reduced. To achieve this, we formulate a new sparse learning model that is able to jointly reduce the domain discrepancy and select informative features for classification. Extensive experiments on real-world data sets demonstrate the effectiveness of the proposed method.",

keywords = "Domain Adaptation, Feature Selection, Sparse Learning Model",

author = "Yuguang Yan and Hanrui Wu and Yuzhong Ye and Chaoyang Bi and Min Lu and Dapeng Liu and Qingyao Wu and Ng, {Michael K.}",

note = "This work was supported by the NSFC (62206111, 62206061, 62272172), China Postdoctoral Science Foundation (2022M721343), Fundamental Research Funds for the Central Universities (21622326), GuangDong Basic and Applied Basic Research Foundation 2023A1515012920, Tip-top Scientific and Technical Innovative Youth Talents of Guangdong Special Support Program (2019TQ05X200), 2022 Tencent Wechat Rhino-Bird Focused Research Program (Tencent WeChat RBFR2022008), HKRGC GRF (12300519, 17201020, 17300021), HKRGC CRF (C1013-21GF, C7004-21GF), Joint NSFC-RGC N-HKU769/21. Publisher Copyright: {\textcopyright} 2023 IEEE.; 39th IEEE International Conference on Data Engineering, ICDE 2023 ; Conference date: 03-04-2023 Through 07-04-2023",

year = "2023",

month = apr,

day = "3",

doi = "10.1109/ICDE55515.2023.00354",

language = "English",

isbn = "9798350322286",

series = "Proceedings - International Conference on Data Engineering",

publisher = "IEEE",

pages = "3855--3856",

booktitle = "Proceedings - 2023 IEEE 39th International Conference on Data Engineering, ICDE 2023",

address = "United States",

url = "https://icde2023.ics.uci.edu/, https://ieeexplore.ieee.org/xpl/conhome/10184508/proceeding",

}

Yan, Y, Wu, H, Ye, Y, Bi, C, Lu, M, Liu, D, Wu, Q & Ng, MK 2023, Transferable Feature Selection for Unsupervised Domain Adaptation: Extended Abstract. in Proceedings - 2023 IEEE 39th International Conference on Data Engineering, ICDE 2023. Proceedings - International Conference on Data Engineering, vol. 2023-April, IEEE, pp. 3855-3856, 39th IEEE International Conference on Data Engineering, ICDE 2023, Anaheim, California, United States, 3/04/23. https://doi.org/10.1109/ICDE55515.2023.00354

Transferable Feature Selection for Unsupervised Domain Adaptation: Extended Abstract. / Yan, Yuguang; Wu, Hanrui; Ye, Yuzhong et al.
Proceedings - 2023 IEEE 39th International Conference on Data Engineering, ICDE 2023. IEEE, 2023. p. 3855-3856 (Proceedings - International Conference on Data Engineering; Vol. 2023-April).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Transferable Feature Selection for Unsupervised Domain Adaptation

T2 - 39th IEEE International Conference on Data Engineering, ICDE 2023

AU - Yan, Yuguang

AU - Wu, Hanrui

AU - Ye, Yuzhong

AU - Bi, Chaoyang

AU - Lu, Min

AU - Liu, Dapeng

AU - Wu, Qingyao

AU - Ng, Michael K.

N1 - This work was supported by the NSFC (62206111, 62206061, 62272172), China Postdoctoral Science Foundation (2022M721343), Fundamental Research Funds for the Central Universities (21622326), GuangDong Basic and Applied Basic Research Foundation 2023A1515012920, Tip-top Scientific and Technical Innovative Youth Talents of Guangdong Special Support Program (2019TQ05X200), 2022 Tencent Wechat Rhino-Bird Focused Research Program (Tencent WeChat RBFR2022008), HKRGC GRF (12300519, 17201020, 17300021), HKRGC CRF (C1013-21GF, C7004-21GF), Joint NSFC-RGC N-HKU769/21. Publisher Copyright: © 2023 IEEE.

PY - 2023/4/3

Y1 - 2023/4/3

N2 - Domain adaptation aims at extracting knowledge from auxiliary source domains to assist the learning task in a target domain. Since the distributions of the source and target domains are different, directly using source data to build a classifier for the target domain may hamper the classification performance on the target data. In this paper, we propose to find a feature subset that is both transferable and discriminative, so that both the domain discrepancy and the classification loss measured on the selected features can be reduced. To achieve this, we formulate a new sparse learning model that is able to jointly reduce the domain discrepancy and select informative features for classification. Extensive experiments on real-world data sets demonstrate the effectiveness of the proposed method.

AB - Domain adaptation aims at extracting knowledge from auxiliary source domains to assist the learning task in a target domain. Since the distributions of the source and target domains are different, directly using source data to build a classifier for the target domain may hamper the classification performance on the target data. In this paper, we propose to find a feature subset that is both transferable and discriminative, so that both the domain discrepancy and the classification loss measured on the selected features can be reduced. To achieve this, we formulate a new sparse learning model that is able to jointly reduce the domain discrepancy and select informative features for classification. Extensive experiments on real-world data sets demonstrate the effectiveness of the proposed method.

KW - Domain Adaptation

KW - Feature Selection

KW - Sparse Learning Model

UR - http://www.scopus.com/inward/record.url?scp=85167662802&partnerID=8YFLogxK

U2 - 10.1109/ICDE55515.2023.00354

DO - 10.1109/ICDE55515.2023.00354

M3 - Conference proceeding

AN - SCOPUS:85167662802

SN - 9798350322286

T3 - Proceedings - International Conference on Data Engineering

SP - 3855

EP - 3856

BT - Proceedings - 2023 IEEE 39th International Conference on Data Engineering, ICDE 2023

PB - IEEE

Y2 - 3 April 2023 through 7 April 2023

ER -

Transferable Feature Selection for Unsupervised Domain Adaptation: Extended Abstract

Abstract

Publication series

Competition

Scopus Subject Areas

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this