Number of the records: 1
Gradient Learning in Networks of Smoothly Spiking Neurons
- 1.
SYSNO ASEP 0318366 Document Type C - Proceedings Paper (int. conf.) R&D Document Type Conference Paper Title Gradient Learning in Networks of Smoothly Spiking Neurons Title Gradientní učení sítí hladce pulzních neuronů Author(s) Šíma, Jiří (UIVT-O) RID, SAI, ORCID Source Title Advances in Neuro-Information Processing. Revised Selected Papers Part II. - Berlin : Springer, 2009 / Köppen M. ; Kasabov N. ; Coghill G. - ISBN 978-3-642-03039-0 Pages s. 179-186 Number of pages 8 s. Action ICONIP 2008. International Conference on Neural Information Processing /15./ Event date 25.11.2008-28.11.2008 VEvent location Auckland Country NZ - New Zealand Event type WRD Language eng - English Country DE - Germany Keywords spiking neuron ; back-propagation ; SpikeProp ; gradient learning Subject RIV IN - Informatics, Computer Science R&D Projects 1ET100300517 GA AV ČR - Academy of Sciences of the Czech Republic (AV ČR) 1M0545 GA MŠMT - Ministry of Education, Youth and Sports (MEYS) CEZ AV0Z10300504 - UIVT-O (2005-2011) UT WOS 000270578200022 EID SCOPUS 70349106368 DOI 10.1007/978-3-642-03040-6_22 Annotation A slightly simplified version of the Spike Response Model SRM0 of a spiking neuron is tailored to gradient learning. In particular, the evolution of spike trains along the weight and delay parameter trajectories is made perfectly smooth. For this model a back-propagation-like learning rule is derived which propagates the error also along the time axis. This approach overcomes the difficulties with the discontinuous-in-time nature of spiking neurons, which encounter previous gradient learning algorithms (e.g. SpikeProp). The new algorithm can naturally cope with multiple spikes and preliminary experiments confirm the smoothness of spike creation/deletion process. Workplace Institute of Computer Science Contact Tereza Šírová, sirova@cs.cas.cz, Tel.: 266 053 800 Year of Publishing 2010
Number of the records: 1