Data-driven modeling from biased small training data using periodic orbits

Kengo Nakai, Yoshitaka Saiki

In this study, we investigate the effect of reservoir computing training data on the reconstruction of chaotic dynamics. Our findings indicate that a training time series comprisin…