Tailoring Echo State Networks for Optimal Learning

Pau Vilimelis Aceituno; Gang Yan; Yang-Yu Liu

doi:10.1016/j.isci.2020.101440

Tailoring Echo State Networks for Optimal Learning

iScience. 2020 Aug 6;23(9):101440. doi: 10.1016/j.isci.2020.101440. eCollection 2020 Sep 25.

Authors

Pau Vilimelis Aceituno^{1

2}, Gang Yan³, Yang-Yu Liu¹

Affiliations

¹ Channing Division of Network Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA 02115, USA.
² Max Planck Institute for Mathematics in the Sciences, 04103 Leipzig, Germany.
³ School of Physics Science and Engineering, Tongji University, 200092 Shanghai, China.

Abstract

As one of the most important paradigms of recurrent neural networks, the echo state network (ESN) has been applied to a wide range of fields, from robotics to medicine, finance, and language processing. A key feature of the ESN paradigm is its reservoir-a directed and weighted network of neurons that projects the input time series into a high-dimensional space where linear regression or classification can be applied. By analyzing the dynamics of the reservoir we show that the ensemble of eigenvalues of the network contributes to the ESN memory capacity. Moreover, we find that adding short loops to the reservoir network can tailor ESN for specific tasks and optimize learning. We validate our findings by applying ESN to forecast both synthetic and real benchmark time series. Our results provide a simple way to design task-specific ESN and offer deep insights for other recurrent neural networks.

Keywords: Artificial Intelligence; Network Algorithm; Network Architecture.

Abstract

Grants and funding