Investigation of simple quantum reservoir computers

A reservoir computer is a neural network model that uses past information to predict a system's future states. Recently, quantum reservoir computers (QRCs) have emerged as a promising research area, with studies suggesting that QRCs may offer greater efficiency compared to classical reservoirs. This paper investigates the applicability of quantum reservoirs, their advantages over classical reservoirs, and presents the results through mathematical analyses and computer simulations. Specifically, we model a quantum harmonic oscillator (the reservoir), with its angular frequency (ω) modulated by the position (𝑥) of a classical noisy damped harmonic oscillator (the system). Our findings, derived from calculations and simulations, demonstrate that quantum reservoirs exhibit superior prediction accuracy compared to classical reservoirs. This improvement is attributable to the infinite-dimensionality of quantum reservoirs, which enables greater information storage. While classical reservoirs can be manipulated to achieve high dimensionality, an infinite-dimensional quantum reservoir retains the capability to store significantly more information. These results open avenues for further exploration of different quantum systems as reservoirs, such as entangled qubit pairs, and their potential applications.