Summary
International Symposium on Nonlinear Theory and its Applications
2017
Session Number:C2L-E-1
Session:
Number:C2L-E-1-1
Statistical Analysis of Chaotic Neuron in the Mutually-Connected Chaotic Search Method
Takafumi Matsuura,
pp.704-707
Publication Date:2017/12/4
Online ISSN:2188-5079
DOI:10.34385/proc.29.C2L-E-1-1
PDF download (389.3KB)
Summary:
To find near-optimal solutions of combinatorial optimization problems, a method which uses mutually- connected chaotic neural network (CNN) has already been proposed. However, it is not so easy to generate feasible solutions of the problems from the CNN, because an out- put of a chaotic neuron takes an analog value. Each neuron generates a complicated spike time-series. In this paper, to decide good solutions of the combinatorial optimization problems from the CNN, we analyzed complexity of the spike time-series from each chaotic neuron by using a statistical measure, such as coefficient of variation (CV) and local variation of interspike intervals (LV), which are frequently used in the field of neuroscience.