Abstract

We scrutinize the use of machine learning, based on reservoir computing, tobuild data-driven effective models of multiscale chaotic systems. We show that,for a wide scale separation, machine learning generates effective models akinto those obtained using multiscale asymptotic techniques and, remarkably,remains effective in predictability also when the scale separation is reduced.We also show that predictability can be improved by hybridizing the reservoirwith an imperfect model.

 

Quick Read (beta)