File Download

There are no files associated with this item.

  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Deep Model Compression and Inference Speedup of Sum-Product Networks on Tensor Trains

TitleDeep Model Compression and Inference Speedup of Sum-Product Networks on Tensor Trains
Authors
KO, CYCHEN, CHe, ZZhang, YBatselier, KWong, N
KeywordsModel compression
sum–product network (SP)
tensor train (TT)
Issue Date2020
PublisherIEEE. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=72
Citation
IEEE Transactions on Neural Networks and Learning Systems, 2020, v. 31 n. 7, p. 2665-2671 How to Cite?
DOI: http://dx.doi.org/10.1109/TNNLS.2019.2928379
AbstractSum-product networks (SPNs) constitute an emerging class of neural networks with clear probabilistic semantics and superior inference speed over other graphical models. This brief reveals an important connection between SPNs and tensor trains (TTs), leading to a new canonical form which we call tensor SPNs (tSPNs). Specifically, we demonstrate the intimate relationship between a valid SPN and a TT. For the first time, through mapping an SPN onto a tSPN and employing specially customized optimization techniques, we demonstrate improvements up to a factor of 100 on both model compression and inference speedup for various data sets with negligible loss in accuracy.
Persistent Identifierhttp://hdl.handle.net/10722/289690
ISSN
2162-237X
2023 Impact Factor: 10.2
2023 SCImago Journal Rankings: 4.170
ISI Accession Number ID
WOS:000546986600035

 

DC FieldValueLanguage
dc.contributor.authorKO, CY-
dc.contributor.authorCHEN, C-
dc.contributor.authorHe, Z-
dc.contributor.authorZhang, Y-
dc.contributor.authorBatselier, K-
dc.contributor.authorWong, N-
dc.date.accessioned2020-10-22T08:16:04Z-
dc.date.available2020-10-22T08:16:04Z-
dc.date.issued2020-
dc.identifier.citationIEEE Transactions on Neural Networks and Learning Systems, 2020, v. 31 n. 7, p. 2665-2671-
dc.identifier.issn2162-237X-
dc.identifier.urihttp://hdl.handle.net/10722/289690-
dc.description.abstractSum-product networks (SPNs) constitute an emerging class of neural networks with clear probabilistic semantics and superior inference speed over other graphical models. This brief reveals an important connection between SPNs and tensor trains (TTs), leading to a new canonical form which we call tensor SPNs (tSPNs). Specifically, we demonstrate the intimate relationship between a valid SPN and a TT. For the first time, through mapping an SPN onto a tSPN and employing specially customized optimization techniques, we demonstrate improvements up to a factor of 100 on both model compression and inference speedup for various data sets with negligible loss in accuracy.-
dc.languageeng-
dc.publisherIEEE. The Journal's web site is located at http://ieeexplore.ieee.org/xpl/RecentIssue.jsp?punumber=72-
dc.relation.ispartofIEEE Transactions on Neural Networks and Learning Systems-
dc.rightsIEEE Transactions on Neural Networks and Learning Systems. Copyright © IEEE.-
dc.rights©20xx IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.-
dc.subjectModel compression-
dc.subjectsum–product network (SP)-
dc.subjecttensor train (TT)-
dc.titleDeep Model Compression and Inference Speedup of Sum-Product Networks on Tensor Trains-
dc.typeArticle-
dc.identifier.emailWong, N: nwong@eee.hku.hk-
dc.identifier.authorityWong, N=rp00190-
dc.description.naturelink_to_subscribed_fulltext-
dc.identifier.doi10.1109/TNNLS.2019.2928379-
dc.identifier.pmid31403446-
dc.identifier.scopuseid_2-s2.0-85088037212-
dc.identifier.hkuros315878-
dc.identifier.volume31-
dc.identifier.issue7-
dc.identifier.spage2665-
dc.identifier.epage2671-
dc.identifier.isiWOS:000546986600035-
dc.publisher.placeUnited States-
dc.identifier.issnl2162-237X-

Export via OAI-PMH Interface in XML Formats

OR

Export to Other Non-XML Formats