Scaling up Echo-State Networks with multiple light scattering

TL;DR

This paper introduces an optical implementation of Echo-State Networks using light scattering media and digital micromirrors, enabling fast, scalable, and power-efficient large networks for time series prediction.

Contribution

It presents a novel optical hardware approach to scale Echo-State Networks using light scattering media, reducing complexity and enabling larger networks.

Findings

Successfully trained binary networks to predict chaotic time series.

Demonstrated fast and power-efficient operation of the optical implementation.

Scalable to very large network sizes.

Abstract

Echo-State Networks and Reservoir Computing have been studied for more than a decade. They provide a simpler yet powerful alternative to Recurrent Neural Networks, every internal weight is fixed and only the last linear layer is trained. They involve many multiplications by dense random matrices. Very large networks are difficult to obtain, as the complexity scales quadratically both in time and memory. Here, we present a novel optical implementation of Echo-State Networks using light-scattering media and a Digital Micromirror Device. As a proof of concept, binary networks have been successfully trained to predict the chaotic Mackey-Glass time series. This new method is fast, power efficient and easily scalable to very large networks.