Show simple item record

dc.contributor.authorSrivastava, Vipinen
dc.contributor.authorSampath, Suchitraen
dc.contributor.authorParker, Daviden
dc.date.accessioned2014-11-07T15:49:48Z
dc.date.available2014-11-07T15:49:48Z
dc.date.issued2014-09-02en
dc.identifier.citationPLoS ONE 9(9): e105619. doi:10.1371/journal.pone.0105619en
dc.identifier.issn1932-6203
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/246322
dc.description.abstractConnectionist models of memory storage have been studied for many years, and aim to provide insight into potential mechanisms of memory storage by the brain. A problem faced by these systems is that as the number of items to be stored increases across a finite set of neurons/synapses, the cumulative changes in synaptic weight eventually lead to a sudden and dramatic loss of the stored information (catastrophic interference, CI) as the previous changes in synaptic weight are effectively lost. This effect does not occur in the brain, where information loss is gradual. Various attempts have been made to overcome the effects of CI, but these generally use schemes that impose restrictions on the system or its inputs rather than allowing the system to intrinsically cope with increasing storage demands. We show here that catastrophic interference occurs as a result of interference among patterns that lead to catastrophic effects when the number of patterns stored exceeds a critical limit. However, when Gram-Schmidt orthogonalization is combined with the Hebb-Hopfield model, the model attains the ability to eliminate CI. This approach differs from previous orthogonalisation schemes used in connectionist networks which essentially reflect sparse coding of the input. Here CI is avoided in a network of a fixed size without setting limits on the rate or number of patterns encoded, and without separating encoding and retrieval, thus offering the advantage of allowing associations between incoming and stored patterns.
dc.description.sponsorshipThe Royal Society; The Leverhulme Foundation (UK); National Initiative of Research in Cognitive Science by the Department of Science and Technology, Government of India.
dc.languageEnglishen
dc.language.isoenen
dc.publisherPLOS
dc.rightsAttribution 2.0 UK: England & Wales*
dc.rights.urihttp://creativecommons.org/licenses/by/2.0/uk/*
dc.titleOvercoming Catastrophic Interference in Connectionist Networks Using Gram-Schmidt Orthogonalizationen
dc.typeArticle
dc.description.versionThis is the final published version. It originally appeared in PLOS ONE at http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0105619.en
prism.publicationDate2014en
prism.publicationNamePLOS ONEen
prism.volume9en
dcterms.dateAccepted2014-07-26en
rioxxterms.versionofrecord10.1371/journal.pone.0105619en
rioxxterms.licenseref.urihttp://www.rioxx.net/licenses/all-rights-reserveden
rioxxterms.licenseref.startdate2014-09-02en
dc.contributor.orcidParker, David [0000-0002-5345-348X]
dc.identifier.eissn1932-6203
rioxxterms.typeJournal Article/Reviewen
pubs.funder-project-idRoyal Society (2013/R1)


Files in this item

Thumbnail
Thumbnail

This item appears in the following Collection(s)

Show simple item record

Attribution 2.0 UK: England & Wales
Except where otherwise noted, this item's licence is described as Attribution 2.0 UK: England & Wales