The Spike Response Model: A framework to predict neuronal spike trains

R Jolivet*, TJ Lewis, W Gerstner

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingAcademicpeer-review

Abstract

We propose a simple method to map a generic threshold model, namely the Spike Response Model, to artificial data of neuronal activity using a minimal amount of a priori information. Here, data are generated by a detailed mathematical model of neuronal activity. The model neuron is driven with in-vivo-like current injected, and we test to which extent it is possible to predict the spike train of the detailed neuron model from that of the Spike Response Model. In particular, we look at the number of spikes correctly predicted within a biologically relevant time window. We find that the Spike Response Model achieves prediction of up to 80% of the spikes with correct timing (+/-2ms). Other characteristics of activity, such as mean rate and coefficient of variation of spike trains, are predicted in the correct range as well.

Original languageEnglish
Title of host publicationARTIFICIAL NEURAL NETWORKS AND NEURAL INFORMATION PROCESSING - ICAN/ICONIP 2003
EditorsO Kaynak, E Alpaydin, E Oja, L Xu
PublisherSpringer Verlag
Pages846-853
Number of pages8
ISBN (Print)3-540-40408-2
DOIs
Publication statusPublished - 2003
EventJoint International Conference on Artificial Neural Networks (ICANN)/International Conference on Neural Information Processing (ICONIP) - Istanbul, Turkey
Duration: 26 Jun 200229 Jun 2002

Publication series

SeriesLecture Notes in Computer Science
Volume2714
ISSN0302-9743

Conference

ConferenceJoint International Conference on Artificial Neural Networks (ICANN)/International Conference on Neural Information Processing (ICONIP)
Country/TerritoryTurkey
CityIstanbul
Period26/06/0229/06/02

Cite this