Quantum Process Learning Through Neural Emulation

TL;DR

This paper introduces a neural network approach to emulate unknown quantum processes by learning from limited measurement data, enabling accurate prediction of measurement outcomes across relevant quantum ensembles.

Contribution

It presents a neural emulation method that constructs an internal representation of quantum processes from few data points, improving quantum process characterization.

Findings

High accuracy in predicting measurement statistics

Effective in quantum computing, photonics, and many-body physics

Requires limited measurement data for training

Abstract

Neural networks are a promising tool for characterizing intermediate-scale quantum devices from limited amounts of measurement data. A challenging problem in this area is to learn the action of an unknown quantum process on an ensemble of physically relevant input states. To tackle this problem, we introduce a neural network that emulates the unknown process by constructing an internal representation of the input ensemble and by mimicking the action of the process at the state representation level. After being trained with measurement data from a few pairs of input/output quantum states, the network becomes able to predict the measurement statistics for all inputs in the ensemble of interest. We show that our model exhibits high accuracy in applications to quantum computing, quantum photonics, and quantum many-body physics.