000885634 001__ 885634
000885634 005__ 20240313103121.0
000885634 0247_ $$2arXiv$$aarXiv:2010.06247
000885634 0247_ $$2Handle$$a2128/25903
000885634 0247_ $$2altmetric$$aaltmetric:92331342
000885634 037__ $$aFZJ-2020-03975
000885634 1001_ $$0P:(DE-Juel1)174585$$aNestler, Sandra$$b0$$ufzj
000885634 245__ $$aUnfolding recurrence by Green's functions for optimized reservoir computing
000885634 260__ $$c2020
000885634 3367_ $$0PUB:(DE-HGF)25$$2PUB:(DE-HGF)$$aPreprint$$bpreprint$$mpreprint$$s1602842442_17894
000885634 3367_ $$2ORCID$$aWORKING_PAPER
000885634 3367_ $$028$$2EndNote$$aElectronic Article
000885634 3367_ $$2DRIVER$$apreprint
000885634 3367_ $$2BibTeX$$aARTICLE
000885634 3367_ $$2DataCite$$aOutput Types/Working Paper
000885634 520__ $$aCortical networks are strongly recurrent, and neurons have intrinsic temporal dynamics. This sets them apart from deep feed-forward networks. Despite the tremendous progress in the application of feed-forward networks and their theoretical understanding, it remains unclear how the interplay of recurrence and non-linearities in recurrent cortical networks contributes to their function. The purpose of this work is to present a solvable recurrent network model that links to feed forward networks. By perturbative methods we transform the time-continuous, recurrent dynamics into an effective feed-forward structure of linear and non-linear temporal kernels. The resulting analytical expressions allow us to build optimal time-series classifiers from random reservoir networks. Firstly, this allows us to optimize not only the readout vectors, but also the input projection, demonstrating a strong potential performance gain. Secondly, the analysis exposes how the second order stimulus statistics is a crucial element that interacts with the non-linearity of the dynamics and boosts performance.
000885634 536__ $$0G:(DE-HGF)POF3-574$$a574 - Theory, modelling and simulation (POF3-574)$$cPOF3-574$$fPOF III$$x0
000885634 536__ $$0G:(EU-Grant)945539$$aHBP SGA3 - Human Brain Project Specific Grant Agreement 3 (945539)$$c945539$$x1
000885634 536__ $$0G:(DE-82)EXS-SF-neuroIC002$$aneuroIC002 - Recurrence and stochasticity for neuro-inspired computation (EXS-SF-neuroIC002)$$cEXS-SF-neuroIC002$$x2
000885634 536__ $$0G:(DE-Juel1)aca_20190115$$aAdvanced Computing Architectures (aca_20190115)$$caca_20190115$$fAdvanced Computing Architectures$$x3
000885634 536__ $$0G:(DE-Juel-1)BMBF-01IS19077A$$aRenormalizedFlows - Transparent Deep Learning with Renormalized Flows (BMBF-01IS19077A)$$cBMBF-01IS19077A$$x4
000885634 536__ $$0G:(DE-Juel-1)PF-JARA-SDS005$$aSDS005 - Towards an integrated data science of complex natural systems (PF-JARA-SDS005)$$cPF-JARA-SDS005$$x5
000885634 536__ $$0G:(DE-Juel1)PHD-NO-GRANT-20170405$$aPhD no Grant - Doktorand ohne besondere Förderung (PHD-NO-GRANT-20170405)$$cPHD-NO-GRANT-20170405$$x6
000885634 588__ $$aDataset connected to arXivarXiv
000885634 7001_ $$0P:(DE-Juel1)171384$$aKeup, Christian$$b1$$ufzj
000885634 7001_ $$0P:(DE-Juel1)156459$$aDahmen, David$$b2$$ufzj
000885634 7001_ $$0P:(DE-Juel1)184621$$aGilson, Matthieu$$b3$$ufzj
000885634 7001_ $$0P:(DE-HGF)0$$aRauhut, Holger$$b4
000885634 7001_ $$0P:(DE-Juel1)144806$$aHelias, Moritz$$b5$$eCorresponding author
000885634 8564_ $$uhttps://juser.fz-juelich.de/record/885634/files/neurips_draft_rev1.pdf$$yOpenAccess
000885634 8564_ $$uhttps://juser.fz-juelich.de/record/885634/files/neurips_draft_rev1.pdf?subformat=pdfa$$xpdfa$$yOpenAccess
000885634 909CO $$ooai:juser.fz-juelich.de:885634$$pdnbdelivery$$pec_fundedresources$$pVDB$$pdriver$$popen_access$$popenaire
000885634 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)174585$$aForschungszentrum Jülich$$b0$$kFZJ
000885634 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)171384$$aForschungszentrum Jülich$$b1$$kFZJ
000885634 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)156459$$aForschungszentrum Jülich$$b2$$kFZJ
000885634 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)184621$$aForschungszentrum Jülich$$b3$$kFZJ
000885634 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)144806$$aForschungszentrum Jülich$$b5$$kFZJ
000885634 9131_ $$0G:(DE-HGF)POF3-574$$1G:(DE-HGF)POF3-570$$2G:(DE-HGF)POF3-500$$3G:(DE-HGF)POF3$$4G:(DE-HGF)POF$$aDE-HGF$$bKey Technologies$$lDecoding the Human Brain$$vTheory, modelling and simulation$$x0
000885634 9141_ $$y2020
000885634 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000885634 9201_ $$0I:(DE-Juel1)INM-6-20090406$$kINM-6$$lComputational and Systems Neuroscience$$x0
000885634 9201_ $$0I:(DE-Juel1)IAS-6-20130828$$kIAS-6$$lTheoretical Neuroscience$$x1
000885634 9201_ $$0I:(DE-Juel1)INM-10-20170113$$kINM-10$$lJara-Institut Brain structure-function relationships$$x2
000885634 9801_ $$aFullTexts
000885634 980__ $$apreprint
000885634 980__ $$aVDB
000885634 980__ $$aUNRESTRICTED
000885634 980__ $$aI:(DE-Juel1)INM-6-20090406
000885634 980__ $$aI:(DE-Juel1)IAS-6-20130828
000885634 980__ $$aI:(DE-Juel1)INM-10-20170113
000885634 981__ $$aI:(DE-Juel1)IAS-6-20130828