TCB Publications - Abstract

Joachim Buhmann and Klaus Schulten. Influence of noise on the behaviour of an autoassociative neural network. In J. S. Denker, editor, Neural Networks for Computing, pp. 71-76. American Institute of Physics Publication, Conference Proceedings 151, 1986.

BUHM86B Recently, we simulated the activity and function of neural networks with neuronal units modelled after their physiological counterparts1,2. Neuronal potentials, single neural spikes and their effects on postsynaptic neurons were taken into account. The neural network studied was endowed with plastic synapses. The synaptic modifications were assumed to follow Hebbian rules, i.e. the synaptic strengths increase if the pre- and postsynaptic cells fire a spike synchronously and decrease if there exists no synchronicity between pre-and postsynaptic spikes. The time scale of the synaptic plasticity was that of mental processes, i.e. a tenth of a second as proposed by v.d. Malsburg3. In this contribution we extend our previous study and include random fluctuations of the neural potentials as observed in electrophysiological recordings4. We will demonstrate that random fluctuations of the membrane potentials raise the sensitivity and performance of the neural network. The fluctuations enable the network to react to weak external stimuli which do not affect networks following deterministic dynamics. We argue that fluctuations and noise in the membrane potential are of functional importance in that they trigger the neural firing if a weak receptor input is presented. The noise regulates the level of arousal. It might be an essential feature of the information processing abilities of neuronal networks and not a mere source of disturbance better to be suppressed. We will demonstrate that the neural network investigated here reproduces the computational abilities of formal associative networks4,5,6.

Download Full Text

The manuscripts available on our site are provided for your personal use only and may not be retransmitted or redistributed without written permissions from the paper's publisher and author. You may not upload any of this site's material to any public server, on-line service, network, or bulletin board without prior written permission from the publisher and author. You may not make copies for any commercial purpose. Reproduction or storage of materials retrieved from this web site is subject to the U.S. Copyright Act of 1976, Title 17 U.S.C.

Download full text: PDF (241.7KB)