Presentation 2006-11-11
An approach toward learning algorithm for Digital Spiking Neuron
Hiroyuki TORIKAI, Toshimichi SAITO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A Digital Spiking Neuron (DSN) can generate a spike-train that can have various sequences of inter-spike intervals (ISIs). The DSN is a wired system of shift registers, and the wiring pattern is a crucial parameter that determines dynamics of the ISI. In this paper, first, we clarify a basic relation between the wiring pattern and basic characteristics of the spike-train. Based on this relation, we propose a learning algorithm that utilizes re-wirings. As a teacher signal, the DSN accepts a sequence of ISIs generated by a teacher-DSN that has an unknown wiring pattern. We then show that the proposed learning algorithm can be an approach toward a fundamental problem: Can the DSN reconstruct the unknown wiring patterns of the teacher-DSN from the ISI sequence?
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Spiking neuron / Pulse-coupled network / shift register generator / Cellular Automata / FPGA / Learning
Paper # NC2006-69
Date of Issue

Conference Information
Committee NC
Conference Date 2006/11/4(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An approach toward learning algorithm for Digital Spiking Neuron
Sub Title (in English)
Keyword(1) Spiking neuron
Keyword(2) Pulse-coupled network
Keyword(3) shift register generator
Keyword(4) Cellular Automata
Keyword(5) FPGA
Keyword(6) Learning
1st Author's Name Hiroyuki TORIKAI
1st Author's Affiliation EECE Dept., Hosei University()
2nd Author's Name Toshimichi SAITO
2nd Author's Affiliation EECE Dept., Hosei University
Date 2006-11-11
Paper # NC2006-69
Volume (vol) vol.106
Number (no) 342
Page pp.pp.-
#Pages 6
Date of Issue