Show simple item record

dc.contributor.author
Qu, Zhongnan
dc.contributor.author
Liu, Cong
dc.contributor.author
Guo, Junfeng
dc.contributor.author
Thiele, Lothar
dc.date.accessioned
2021-02-10T07:03:51Z
dc.date.available
2021-01-04T15:25:30Z
dc.date.available
2021-02-10T07:03:51Z
dc.date.issued
2020-07-06
dc.identifier.uri
http://hdl.handle.net/20.500.11850/459131
dc.description.abstract
Emerging edge intelligence applications require the server to continuously retrain and update deep neural networks deployed on remote edge nodes in order to leverage newly collected data samples. Unfortunately, it may be impossible in practice to continuously send fully updated weights to these edge nodes due to the highly constrained communication resource. In this paper, we propose the weight-wise deep partial updating paradigm, which smartly selects only a subset of weights to update at each server-to-edge communication round, while achieving a similar performance compared to full updating. Our method is established through analytically upper-bounding the loss difference between partial updating and full updating, and only updates the weights which make the largest contributions to the upper bound. Extensive experimental results demonstrate the efficacy of our partial updating methodology which achieves a high inference accuracy while updating a rather small number of weights.
en_US
dc.language.iso
en
en_US
dc.publisher
Cornell University
en_US
dc.title
Deep Partial Updating
en_US
dc.type
Working Paper
ethz.journal.title
arXiv
ethz.pages.start
2007.03071
en_US
ethz.size
20 p.
en_US
ethz.identifier.arxiv
2007.03071
ethz.publication.place
Ithaca, NY
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02140 - Dep. Inf.technologie und Elektrotechnik / Dep. of Inform.Technol. Electrical Eng.::02640 - Inst. f. Technische Informatik und Komm. / Computer Eng. and Networks Lab.::03429 - Thiele, Lothar (emeritus) / Thiele, Lothar (emeritus)
en_US
ethz.leitzahl.certified
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02140 - Dep. Inf.technologie und Elektrotechnik / Dep. of Inform.Technol. Electrical Eng.::02640 - Inst. f. Technische Informatik und Komm. / Computer Eng. and Networks Lab.::03429 - Thiele, Lothar (emeritus) / Thiele, Lothar (emeritus)
en_US
ethz.date.deposited
2021-01-04T15:25:37Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Metadata only
en_US
ethz.rosetta.installDate
2021-02-10T07:04:09Z
ethz.rosetta.lastUpdated
2023-02-06T21:25:25Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Deep%20Partial%20Updating&rft.jtitle=arXiv&rft.date=2020-07-06&rft.spage=2007.03071&rft.au=Qu,%20Zhongnan&Liu,%20Cong&Guo,%20Junfeng&Thiele,%20Lothar&rft.genre=preprint&
 Search print copy at ETH Library

Files in this item

FilesSizeFormatOpen in viewer

There are no files associated with this item.

Publication type

Show simple item record