On the experimental feasibility of quantum state reconstruction via machine learning

TL;DR

This paper evaluates the resource requirements and performance of machine learning methods for quantum state reconstruction, including experimental implementation on IBM Q, focusing on pure states of small systems and low-count regimes.

Contribution

It provides a detailed analysis of resource scaling, compares machine learning with traditional methods, and demonstrates practical implementation on a real quantum computer.

Findings

Machine learning methods scale favorably for small systems.

Performance degrades in low-count regimes but remains promising.

Experimental validation on IBM Q shows competitive results.

Abstract

We determine the resource scaling of machine learning-based quantum state reconstruction methods, in terms of inference and training, for systems of up to four qubits when constrained to pure states. Further, we examine system performance in the low-count regime, likely to be encountered in the tomography of high-dimensional systems. Finally, we implement our quantum state reconstruction method on an IBM Q quantum computer, and compare against both unconstrained and constrained MLE state reconstruction.