Demonstration of a bosonic quantum classifier with data re-uploading

TL;DR

This paper demonstrates a bosonic quantum classifier using data re-uploading in silicon optical circuits, achieving high classification accuracy and paving the way for advanced quantum machine learning with multi-photon states.

Contribution

It introduces a new quantum classifier for bosonic systems applying data re-uploading, demonstrated with silicon optical circuits, and establishes a theoretical framework for such classifiers.

Findings

Achieved approximately 94% classification accuracy in proof-of-principle experiments.

Developed a programmable optical quantum circuit with an interferometer.

Proposed extension to multi-photon and entangled states for future quantum classifiers.

Abstract

In a single qubit system, a universal quantum classifier can be realised using the data-reuploading technique. In this study, we propose a new quantum classifier applying this technique to bosonic systems and successfully demonstrated it using silicon optical integrated quantum circuits. We established a theory of quantum machine learning algorithm applicable to bosonic systems and implemented a programmable optical circuit combined with an interferometer. Learning and classification using part of the implemented optical quantum circuit with uncorrelated two-photons resulted in a classification with a reproduction rate of approximately 94\% in the proof of principle experiment. As this method can be applied to arbitrary two-mod N-photon system, further development of optical quantum classifiers, such as extensions to quantum entangled and multi-photon states, is expected in the future.