Summary
International Symposium on Nonlinear Theory and its Applications
2017
Session Number:A2L-E
Session:
Number:A2L-E-2
Nonlinear Dynamics of Memristive Networks and its Application to Reservoir Computing
Gouhei Tanaka, Ryosho Nakane, Toshiyuki Yamane, Seiji Takeda, Daiju Nakano, Shigeru Nakagawa, Akira Hirose,
pp.182-185
Publication Date:2017/12/4
Online ISSN:2188-5079
Summary:
Reservoir computing is one of the potent computational frameworks suitable for seuqential data processing. Not only recurrent neural networks but also other physical systems and devices are available to construct a reservoir computing system. In this study, we focus on memristive networks consisting of coupled memristors for achieving physical reservoir computing. First, we present a mathematical model of memristive network circuits with any architecture and investigate its nonlinear dynamics. The dynamical response to input sequential data is also examined. Next, we deal with the problem of how to design memristive networks for better computational performance in a reservoir computing framework. Finally, we make a discussion toward device implementaion of our system.