Zhurnal Vychislitel'noi Matematiki i Matematicheskoi Fiziki
RUS  ENG    JOURNALS   PEOPLE   ORGANISATIONS   CONFERENCES   SEMINARS   VIDEO LIBRARY   PACKAGE AMSBIB  
General information
Latest issue
Archive
Impact factor

Search papers
Search references

RSS
Latest issue
Current issues
Archive issues
What is RSS


Zh. Vychisl. Mat. Mat. Fiz.:
Year:
Volume:
Issue:
Page:
Find




Personal entry:
Login:
Password:
Save password
Enter
Forgotten password?
Register

Zhurnal Vychislitel'noi Matematiki i Matematicheskoi Fiziki, 2021, Volume 61, Number 5, Pages 800–812
DOI: https://doi.org/10.31857/S0044466921050100 (Mi zvmmf11239) 		 
This article is cited in 5 scientific papers (total in 5 papers)

General numerical methods

Reduced-order modeling of deep neural networks

J. V. Gusaka, T. K. Daulbaeva, I. V. Oseledetsab, E. S. Ponomareva, A. S. Cichockia

a Skolkovo Institute of Science and Technology, Moscow, Russia
b Marchuk Institute of Numerical Mathematics of the Russian Academy of Sciences, Moscow
Citations (5)
DOI: https://doi.org/10.31857/S0044466921050100
Abstract: We introduce a new method for speeding up the inference of deep neural networks. It is somewhat inspired by the reduced-order modeling techniques for dynamical systems. The cornerstone of the proposed method is the maximum volume algorithm. We demonstrate efficiency on neural networks pre-trained on different datasets. We show that in many practical cases it is possible to replace convolutional layers with much smaller fully-connected layers with a relatively small drop in accuracy.
Key words: acceleration of neural networks, MaxVol, machine learning, component analysis.
Funding agency Grant number
Russian Foundation for Basic Research 19-31-90172
20-31-90127
Ministry of Education and Science of the Russian Federation 14.756.31.0001
This study was supported by RFBR, project nos. 19-31-90172 and 20-31-90127 (algorithm) and by the Ministry of Education and Science of the Russian Federation (grant 14.756.31.0001) (experiments).
Received: 24.12.2020
Revised: 24.12.2020
Accepted: 14.01.2021
English version:
Computational Mathematics and Mathematical Physics, 2021, Volume 61, Issue 5, Pages 774–785
DOI: https://doi.org/10.1134/S0965542521050109
Bibliographic databases:
Document Type: Article
UDC: 519.65
Language: Russian
Citation: J. V. Gusak, T. K. Daulbaev, I. V. Oseledets, E. S. Ponomarev, A. S. Cichocki, “Reduced-order modeling of deep neural networks”, Zh. Vychisl. Mat. Mat. Fiz., 61:5 (2021), 800–812; Comput. Math. Math. Phys., 61:5 (2021), 774–785
Citation in format AMSBIB
\Bibitem{GusDauOse21}
\by J.~V.~Gusak, T.~K.~Daulbaev, I.~V.~Oseledets, E.~S.~Ponomarev, A.~S.~Cichocki
\paper Reduced-order modeling of deep neural networks
\jour Zh. Vychisl. Mat. Mat. Fiz.
\yr 2021
\vol 61
\issue 5
\pages 800--812
\mathnet{http://mi.mathnet.ru/zvmmf11239}
\crossref{https://doi.org/10.31857/S0044466921050100}
\elib{https://elibrary.ru/item.asp?id=45633450}
\transl
\jour Comput. Math. Math. Phys.
\yr 2021
\vol 61
\issue 5
\pages 774--785
\crossref{https://doi.org/10.1134/S0965542521050109}
\isi{https://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcApp=Publons&SrcAuth=Publons_CEL&DestLinkType=FullRecord&DestApp=WOS_CPL&KeyUT=000668966500009}
\scopus{https://www.scopus.com/record/display.url?origin=inward&eid=2-s2.0-85109105535}
Linking options:
  • https://www.mathnet.ru/eng/zvmmf11239
  • https://www.mathnet.ru/eng/zvmmf/v61/i5/p800
    • This publication is cited in the following 5 articles:
    Citing articles in Google Scholar: Russian citations, English citations
    Related articles in Google Scholar: Russian articles, English articles
    		Журнал вычислительной математики и математической физики Computational Mathematics and Mathematical Physics
     
      Contact us:
    		  Terms of Use  Registration to the website  Logotypes © Steklov Mathematical Institute RAS, 2024