Reservoir Computing: Machine Learning Meets Nonlinear Dynamics

Abstract:

Reservoir computing has recently been exploited to solve a variety of challenging problems in complex nonlinear dynamical systems. The speaker will review some recent works from his group in this area: predicting tipping point and critical transitions, digital twins of nonlinear dynamical systems, parameter and trajectory tracking, and associative memory for complex dynamical patterns. Some open questions will be discussed.

Collaborators: Shirin Panahi, Ling-Wei Kong, Zheng-Meng Zhai, and Mohammadamin Moradi