search for




 

https://doi.org/10.3938/NPSM.69.874
Study of the Signal Response Domain in a Spiking Neural Network Model
New Phys.: Sae Mulli 2019; 69: 874~881
Published online August 30, 2019;  https://doi.org/10.3938/NPSM.69.874
© 2019 New Physics: Sae Mulli.

Myoung Won CHO*

Department of Global Medical Science, Sungshin Women's University, Seoul 01133, Korea
Correspondence to: mwcho@sungshin.ac.kr
Received June 25, 2019; Revised July 8, 2019; Accepted August 8, 2019.
cc This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Compared with the artificial neural network model, a biological neural network or a spiking neural network model has the distinctive feature that the difference between input spiking timings plays the key role in the determination of an output neuron's reactivity. The input vector space of a spiking neural network is also defined from the relative spiking timings. We here explore what response domain a spiking neural network can have depending on its structure. While the response domain of an output neuron, connected directly from input neurons, in an artificial neural network has a deterministic surface with the shape of a linear plane, that in the spiking neural network has a deterministic surface(s) with the shape of a closed or a curved plane. We suggest and demonstrate how a spiking neural network can have a variety of forms for response domains through not only a multilayer structure but also a complex structure having different connections numbers per signaling path. These results can be applied to understanding or designing a biological neural network with a perceptive function.
PACS numbers: 87.19.lj, 87.19.ll
Keywords: Neural network dynamics, Spiking neural network model


November 2019, 69 (11)
  • Scopus
  • CrossMark