Summary
International Symposium on Nonlinear Theory and its Applications
2017
Session Number:A3L-C
Session:
Number:A3L-C-5
Assimilating nonlinear dynamics with FORCE-learning : A perspective from chaotic synchronization
Hiromichi Suetani,
pp.268-270
Publication Date:2017/12/4
Online ISSN:2188-5079
DOI:10.34385/proc.29.A3L-C-5
PDF download (669KB)
Summary:
We propose an approach for modeling dynamical systems using FORCE-learning, a version of reservoir computing (RC) framework. In this approach, a direct coupling between FORCE-learning systems and the target dynamical system is employed, which enables us to treat the problem of the system identification with terms of synchronization phenomena. Several examples including limit cycle oscillators and chaotic systems are tested as demonstrations and we investigate how our approach is useful for modeling general nonlinear dynamics.