A Biologically Inspired Training Algorithm for Spiking Neural Networks

John Wade, Liam McDaid, JA Santos, Heather Sayers

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The work presented in this paper merges the Bienenstock-Cooper-Munro (BCM) learning rule with the Spike Timing Dependant Plasticity (STDP) rule to develop a training algorithm for a multi layer Spiking Neural Network (SNN), stimulated using spike trains. The BCM rule is utilised to modulate the height of the plasticity window, associated with STDP, as a function of the activity of the postsynaptic neurons, and in doing so introduces a correlation between the activity of the postsynaptic neurons and their associated weights. The induced correlation uses the activity of postsynaptic neurons to stabilise the weight values across a multi-layer network causing convergence during training. The training algorithm also includes both exhibitory and inhibitory facilitating dynamic synapses that create a frequency filtering mechanism allowing the information presented to the network to be routed to different neurons. A variable neuron threshold level simulates the refractory period.
LanguageEnglish
Title of host publicationUnknown Host Publication
Pages7-12
Number of pages6
Publication statusPublished - 13 Sep 2007
EventIrish Signals and Systems Conference - Derry, Ireland
Duration: 13 Sep 2007 → …

Conference

ConferenceIrish Signals and Systems Conference
Period13/09/07 → …

Fingerprint

Neurons
Weights and Measures
Synapses
Learning

Cite this

@inproceedings{b41b74e886de438eb2a9e2deabe199a6,
title = "A Biologically Inspired Training Algorithm for Spiking Neural Networks",
abstract = "The work presented in this paper merges the Bienenstock-Cooper-Munro (BCM) learning rule with the Spike Timing Dependant Plasticity (STDP) rule to develop a training algorithm for a multi layer Spiking Neural Network (SNN), stimulated using spike trains. The BCM rule is utilised to modulate the height of the plasticity window, associated with STDP, as a function of the activity of the postsynaptic neurons, and in doing so introduces a correlation between the activity of the postsynaptic neurons and their associated weights. The induced correlation uses the activity of postsynaptic neurons to stabilise the weight values across a multi-layer network causing convergence during training. The training algorithm also includes both exhibitory and inhibitory facilitating dynamic synapses that create a frequency filtering mechanism allowing the information presented to the network to be routed to different neurons. A variable neuron threshold level simulates the refractory period.",
author = "John Wade and Liam McDaid and JA Santos and Heather Sayers",
year = "2007",
month = "9",
day = "13",
language = "English",
pages = "7--12",
booktitle = "Unknown Host Publication",

}

Wade, J, McDaid, L, Santos, JA & Sayers, H 2007, A Biologically Inspired Training Algorithm for Spiking Neural Networks. in Unknown Host Publication. pp. 7-12, Irish Signals and Systems Conference, 13/09/07.

A Biologically Inspired Training Algorithm for Spiking Neural Networks. / Wade, John; McDaid, Liam; Santos, JA; Sayers, Heather.

Unknown Host Publication. 2007. p. 7-12.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - A Biologically Inspired Training Algorithm for Spiking Neural Networks

AU - Wade, John

AU - McDaid, Liam

AU - Santos, JA

AU - Sayers, Heather

PY - 2007/9/13

Y1 - 2007/9/13

N2 - The work presented in this paper merges the Bienenstock-Cooper-Munro (BCM) learning rule with the Spike Timing Dependant Plasticity (STDP) rule to develop a training algorithm for a multi layer Spiking Neural Network (SNN), stimulated using spike trains. The BCM rule is utilised to modulate the height of the plasticity window, associated with STDP, as a function of the activity of the postsynaptic neurons, and in doing so introduces a correlation between the activity of the postsynaptic neurons and their associated weights. The induced correlation uses the activity of postsynaptic neurons to stabilise the weight values across a multi-layer network causing convergence during training. The training algorithm also includes both exhibitory and inhibitory facilitating dynamic synapses that create a frequency filtering mechanism allowing the information presented to the network to be routed to different neurons. A variable neuron threshold level simulates the refractory period.

AB - The work presented in this paper merges the Bienenstock-Cooper-Munro (BCM) learning rule with the Spike Timing Dependant Plasticity (STDP) rule to develop a training algorithm for a multi layer Spiking Neural Network (SNN), stimulated using spike trains. The BCM rule is utilised to modulate the height of the plasticity window, associated with STDP, as a function of the activity of the postsynaptic neurons, and in doing so introduces a correlation between the activity of the postsynaptic neurons and their associated weights. The induced correlation uses the activity of postsynaptic neurons to stabilise the weight values across a multi-layer network causing convergence during training. The training algorithm also includes both exhibitory and inhibitory facilitating dynamic synapses that create a frequency filtering mechanism allowing the information presented to the network to be routed to different neurons. A variable neuron threshold level simulates the refractory period.

M3 - Conference contribution

SP - 7

EP - 12

BT - Unknown Host Publication

ER -