Efficient Characterization of Quantum Evolutions via a Recommender System

TL;DR

This paper introduces a matrix factorization-based recommender system to efficiently characterize quantum evolutions, including correlations and fidelity, in systems up to 10 qubits, offering computational advantages and noise discrimination.

Contribution

It presents a novel application of recommender systems for quantum evolution characterization, enabling efficient analysis of quantum correlations and fidelity in larger systems.

Findings

RS can distinguish noisy from clean quantum correlation data

Significant computational savings in quantum discord database construction

Effective characterization of nonunitary evolutions and discord phase space

Abstract

We demonstrate characterizing quantum evolutions via matrix factorization algorithm, a particular type of the recommender system (RS). A system undergoing a quantum evolution can be characterized in several ways. Here we choose (i) quantum correlations quantified by measures such as entropy, negativity, or discord, and (ii) state-fidelity. Using quantum registers with up to 10 qubits, we demonstrate that an RS can efficiently characterize both unitary and nonunitary evolutions. After carrying out a detailed performance analysis of the RS in two qubits, we show that it can be used to distinguish a clean database of quantum correlations from a noisy or a fake one. Moreover, we find that the RS brings about a significant computational advantage for building a large database of quantum discord, for which no simple closed-form expression exists. Also, RS can efficiently characterize systems undergoing nonunitary evolutions in terms of quantum discord reduction as well as state-fidelity. Finally, we utilize RS for the construction of discord phase space in a nonlinear quantum system.