Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • 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
A Theory of Sequence Indexing and Working Memory in Recurrent Neural Networks
Redwood Center for Theoretical Neuroscience, University of California, Berkeley.
Luleå University of Technology, Department of Computer Science, Electrical and Space Engineering, Computer Science.
Redwood Center for Theoretical Neuroscience, University of California, Berkeley.
2018 (English)In: Neural Computation, ISSN 0899-7667, E-ISSN 1530-888XArticle in journal (Refereed) Epub ahead of print
Abstract [en]

To accommodate structured approaches of neural computation, we propose a class of recurrent neural networks for indexing and storing sequences of symbols or analog data vectors. These networks with randomized input weights and orthogonal recurrent weights implement coding principles previously described in vector symbolic architectures (VSA) and leverage properties of reservoir computing. In general, the storage in reservoir computing is lossy, and cross-talk noise limits the retrieval accuracy and information capacity. A novel theory to optimize memory performance in such networks is presented and compared with simulation experiments. The theory describes linear readout of analog data and readout with winner-take-all error correction of symbolic data as proposed in VSA models. We find that diverse VSA models from the literature have universal performance properties, which are superior to what previous analyses predicted. Further, we propose novel VSA models with the statistically optimal Wiener filter in the readout that exhibit much higher information capacity, in particular for storing analog data. The theory we present also applies to memory buffers, networks with gradual forgetting, which can operate on infinite data streams without memory overflow. Interestingly, we find that different forgetting mechanisms, such as attenuating recurrent weights or neural nonlinearities, produce very similar behavior if the forgetting time constants are aligned. Such models exhibit extensive capacity when their forgetting time constant is optimized for given noise conditions and network size. These results enable the design of new types of VSA models for the online processing of data streams.

Place, publisher, year, edition, pages
MIT Press, 2018.
National Category
Computer Sciences
Research subject
Dependable Communication and Computation Systems
Identifiers
URN: urn:nbn:se:ltu:diva-68365DOI: 10.1162/neco_a_01084PubMedID: 29652585OAI: oai:DiVA.org:ltu-68365DiVA, id: diva2:1197913
Available from: 2018-04-16 Created: 2018-04-16 Last updated: 2018-04-16

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textPubMed

Search in DiVA

By author/editor
Kleyko, Denis
By organisation
Computer Science
In the same journal
Neural Computation
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar

doi
pubmed
urn-nbn

Altmetric score

doi
pubmed
urn-nbn
Total: 12 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • 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