Summary
International Symposium on Nonlinear Theory and Its Applications
2016
Session Number:A4L-G
Session:
Number:A4L-G-2
Desynchronization Induced by Highly Heterogeneous Strengths of Inhibitory Current on Electronic Neuron
Hisashi Kada, Isao Tokuda,
pp.-
Publication Date:2016/11/27
Online ISSN:2188-5079
Summary:
In neural networks research, neural information is assumed to be generally represented by a spatio-temporal firing pattern of neurons. When the neurons generate synchronized spike trains with each other, their capability of information representation is severely limited, because all the neurons represent merely a single temporal pattern. On the other hand, when the neuronal activities become desynchronization, the network may exhibit much more variations of their firing patterns, increasing the information processing capability. The aim of the our study is to consider a way how to avoid such synchronized neuronal firings that degrade the information processing capability. First, as a typical situation to induce neuronal synchrony, we introduce common inputs to a feed-forward neural network. Then, to suppress the synchrony, heterogeneity is introduced to inhibitory inputs. Using an electric circuit, we verify that the heterogeneous inputs indeed give rise to asynchronous firings among the electronic neurons.