A machine learning based approach to the identification of spectral densities in quantum open systems

TL;DR

This paper introduces a machine learning method utilizing neural networks to classify and estimate parameters of spectral densities in quantum open systems, specifically applied to the spin-boson model, enhancing quantum noise spectroscopy capabilities.

Contribution

It presents a novel neural network approach for classifying spectral density types and estimating their parameters in quantum systems, improving environmental characterization techniques.

Findings

High accuracy in spectral density classification

Robust estimation of spectral density parameters

Potential for advancing quantum noise spectroscopy

Abstract

We present a machine learning-based approach for characterising the environment that affects the dynamics of an open quantum system. We focus on the case of an exactly solvable spin-boson model, where the system-environment interaction, whose strength is encoded in the spectral density, induces pure dephasing. By using artificial neural networks trained on the Fourier-transformed time evolution of some observables of the system, we perform both classification -- distinguishing sub-Ohmic, Ohmic, and super-Ohmic spectral densities -- and regression -- thus estimating key parameters of the spectral density function, when the latter is expressed through a power law. Our results demonstrate high classification accuracy and robust parameter estimation, highlighting the potential of machine learning as a powerful tool for probing environmental features in quantum systems and advancing quantum noise spectroscopy.