File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Domain-attention Conditional Wasserstein Distance for Multi-source Domain Adaptation

TitleDomain-attention Conditional Wasserstein Distance for Multi-source Domain Adaptation
Authors
Keywordsattention
Domain adaptation
multiple sources
optimal transport
Issue Date2020
PublisherAssociation for Computing Machinery. The Journal's web site is located at https://tist.acm.org/
Citation
ACM Transactions on Intelligent Systems and Technology, 2020, v. 11 n. 4, p. article no. 44 How to Cite?
AbstractMulti-source domain adaptation has received considerable attention due to its effectiveness of leveraging the knowledge from multiple related sources with different distributions to enhance the learning performance. One of the fundamental challenges in multi-source domain adaptation is how to determine the amount of knowledge transferred from each source domain to the target domain. To address this issue, we propose a new algorithm, called Domain-attention Conditional Wasserstein Distance (DCWD), to learn transferred weights for evaluating the relatedness across the source and target domains. In DCWD, we design a new conditional Wasserstein distance objective function by taking the label information into consideration to measure the distance between a given source domain and the target domain. We also develop an attention scheme to compute the transferred weights of different source domains based on their conditional Wasserstein distances to the target domain. After that, the transferred weights can be used to reweight the source data to determine their importance in knowledge transfer. We conduct comprehensive experiments on several real-world data sets, and the results demonstrate the effectiveness and efficiency of the proposed method.
Persistent Identifierhttp://hdl.handle.net/10722/293375
ISSN
2019 Impact Factor: 2.672
2015 SCImago Journal Rankings: 1.516
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorWu, H-
dc.contributor.authorYan, Y-
dc.contributor.authorNg, MK-
dc.contributor.authorWu, Q-
dc.date.accessioned2020-11-23T08:15:50Z-
dc.date.available2020-11-23T08:15:50Z-
dc.date.issued2020-
dc.identifier.citationACM Transactions on Intelligent Systems and Technology, 2020, v. 11 n. 4, p. article no. 44-
dc.identifier.issn2157-6904-
dc.identifier.urihttp://hdl.handle.net/10722/293375-
dc.description.abstractMulti-source domain adaptation has received considerable attention due to its effectiveness of leveraging the knowledge from multiple related sources with different distributions to enhance the learning performance. One of the fundamental challenges in multi-source domain adaptation is how to determine the amount of knowledge transferred from each source domain to the target domain. To address this issue, we propose a new algorithm, called Domain-attention Conditional Wasserstein Distance (DCWD), to learn transferred weights for evaluating the relatedness across the source and target domains. In DCWD, we design a new conditional Wasserstein distance objective function by taking the label information into consideration to measure the distance between a given source domain and the target domain. We also develop an attention scheme to compute the transferred weights of different source domains based on their conditional Wasserstein distances to the target domain. After that, the transferred weights can be used to reweight the source data to determine their importance in knowledge transfer. We conduct comprehensive experiments on several real-world data sets, and the results demonstrate the effectiveness and efficiency of the proposed method.-
dc.languageeng-
dc.publisherAssociation for Computing Machinery. The Journal's web site is located at https://tist.acm.org/-
dc.relation.ispartofACM Transactions on Intelligent Systems and Technology-
dc.rightsACM Transactions on Intelligent Systems and Technology. Copyright © Association for Computing Machinery.-
dc.rights©ACM, YYYY. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in PUBLICATION, {VOL#, ISS#, (DATE)} http://doi.acm.org/10.1145/nnnnnn.nnnnnn-
dc.subjectattention-
dc.subjectDomain adaptation-
dc.subjectmultiple sources-
dc.subjectoptimal transport-
dc.titleDomain-attention Conditional Wasserstein Distance for Multi-source Domain Adaptation-
dc.typeArticle-
dc.identifier.emailYan, Y: ygyan@hku.hk-
dc.identifier.emailNg, MK: michael.ng@hku.hk-
dc.identifier.authorityNg, MK=rp02578-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1145/3391229-
dc.identifier.scopuseid_2-s2.0-85089280894-
dc.identifier.hkuros319012-
dc.identifier.volume11-
dc.identifier.issue4-
dc.identifier.spagearticle no. 44-
dc.identifier.epagearticle no. 44-
dc.identifier.isiWOS:000583127700009-
dc.publisher.placeUnited States-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats