Quantum semi-supervised generative adversarial network for enhanced data classification

TL;DR

This paper introduces a quantum semi-supervised GAN that combines a quantum generator with a classical classifier to improve data classification accuracy, leveraging quantum expressibility without requiring pure quantum states.

Contribution

It presents a novel quantum GAN architecture with a quantum generator and classical discriminator, demonstrating enhanced expressibility and classification performance in simulations.

Findings

Quantum generator acts as a stronger adversary than classical counterparts.

The system achieves higher classification accuracy in simulations.

No need for data loading or pure quantum states in the generator.

Abstract

In this paper, we propose the quantum semi-supervised generative adversarial network (qSGAN). The system is composed of a quantum generator and a classical discriminator/classifier (D/C). The goal is to train both the generator and the D/C, so that the latter may get a high classification accuracy for a given dataset. The generator needs neither any data loading nor to generate a pure quantum state, while it is expected to serve as a stronger adversary than a classical one thanks to its rich expressibility. These advantages are demonstrated in a numerical simulation.