Full metadata record
DC FieldValueLanguage
dc.contributor.authorGlüge, Stefan-
dc.contributor.authorBöck, Ronald-
dc.contributor.authorPalm, Günther-
dc.contributor.authorWendemuth, Andreas-
dc.date.accessioned2019-02-13T14:05:20Z-
dc.date.available2019-02-13T14:05:20Z-
dc.date.issued2014-
dc.identifier.issn0925-2312de_CH
dc.identifier.issn1872-8286de_CH
dc.identifier.urihttps://digitalcollection.zhaw.ch/handle/11475/15300-
dc.description.abstractIn general, recurrent neural networks have difficulties in learning long-term dependencies. The segmented-memory recurrent neural network (SMRNN) architecture together with the extended real-time recurrent learning (eRTRL) algorithm was proposed to circumvent this problem. Due to its computational complexity eRTRL becomes impractical with increasing network size. Therefore, we introduce the less complex extended backpropagation through time (eBPTT) for SMRNN together with a layer-local unsupervised pre-training procedure. A comparison on the information latching problem showed that eRTRL is better able to handle the latching of information over longer periods of time, even though eBPTT guaranteed a better generalisation when training was successful. Further, pre-training significantly improved the ability to learn long-term dependencies with eBPTT. Therefore, the proposed eBPTT algorithm is suited for tasks that require big networks where eRTRL is impractical. The pre-training procedure itself is independent of the supervised learning algorithm and can improve learning in SMRNN in general.de_CH
dc.language.isoende_CH
dc.publisherElsevierde_CH
dc.relation.ispartofNeurocomputingde_CH
dc.rightsLicence according to publishing contractde_CH
dc.subject.ddc004: Informatikde_CH
dc.titleLearning long-term dependencies in segmented-memory recurrent neural networks with backpropagation of errorde_CH
dc.typeBeitrag in wissenschaftlicher Zeitschriftde_CH
dcterms.typeTextde_CH
zhaw.departementLife Sciences und Facility Managementde_CH
zhaw.organisationalunitInstitut für Angewandte Simulation (IAS)de_CH
dc.identifier.doi10.1016/j.neucom.2013.11.043de_CH
zhaw.funding.euNode_CH
zhaw.originated.zhawYesde_CH
zhaw.pages.end64de_CH
zhaw.pages.start54de_CH
zhaw.publication.statuspublishedVersionde_CH
zhaw.volume141de_CH
zhaw.publication.reviewPeer review (Publikation)de_CH
zhaw.webfeedBio-Inspired Modelling and Learning Systemsde_CH
Appears in Collections:Publikationen Life Sciences und Facility Management

Files in This Item:
There are no files associated with this item.


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.