kth.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Use of Deterministic Transforms to Design Weight Matrices of a Neural Network
KTH, School of Electrical Engineering and Computer Science (EECS).
KTH, School of Electrical Engineering and Computer Science (EECS), Intelligent systems, Information Science and Engineering.ORCID iD: 0000-0003-4406-536x
KTH, School of Electrical Engineering and Computer Science (EECS), Intelligent systems, Information Science and Engineering.ORCID iD: 0000-0002-8534-7622
KTH, School of Electrical Engineering and Computer Science (EECS), Intelligent systems, Information Science and Engineering.ORCID iD: 0000-0003-2638-6047
2021 (English)In: 29th European Signal Processing Conference (EUSIPCO 2021), European Association for Signal, Speech and Image Processing (EURASIP) , 2021, p. 1366-1370Conference paper, Published paper (Refereed)
Abstract [en]

Self size-estimating feedforward network (SSFN) is a feedforward multilayer network. For the existing SSFN, a part of each weight matrix is trained using a layer-wise convex optimization approach (a supervised training), while the other part is chosen as a random matrix instance (an unsupervised training). In this article, the use of deterministic transforms instead of random matrix instances for the SSFN weight matrices is explored. The use of deterministic transforms provides a reduction in computational complexity. The use of several deterministic transforms is investigated, such as discrete cosine transform, Hadamard transform, Hartley transform, and wavelet transforms. The choice of a deterministic transform among a set of transforms is made in an unsupervised manner. To this end, two methods based on features' statistical parameters are developed. The proposed methods help to design a neural net where deterministic transforms can vary across its layers' weight matrices. The effectiveness of the proposed approach vis-a-vis the SSFN is illustrated for object classification tasks using several benchmark datasets.

Place, publisher, year, edition, pages
European Association for Signal, Speech and Image Processing (EURASIP) , 2021. p. 1366-1370
Series
European Signal Processing Conference, ISSN 2076-1465
Keywords [en]
Multilayer neural network, deterministic transforms, weight matrices
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:kth:diva-311283DOI: 10.23919/EUSIPCO54536.2021.9616182ISI: 000764066600272Scopus ID: 2-s2.0-85123210853OAI: oai:DiVA.org:kth-311283DiVA, id: diva2:1653592
Conference
29th European Signal Processing Conference, EUSIPCO 2021, Dublin, 23 August 2021 through 27 August 2021
Note

QC 20220422

Part of proceedings: ISBN 978-9-0827-9706-0

Available from: 2022-04-22 Created: 2022-04-22 Last updated: 2022-06-25Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textScopus

Authority records

Liang, XinyueJavid, Alireza M.Chatterjee, Saikat

Search in DiVA

By author/editor
Jurado, Pol GrauLiang, XinyueJavid, Alireza M.Chatterjee, Saikat
By organisation
School of Electrical Engineering and Computer Science (EECS)Information Science and Engineering
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 20 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf