Quantum Wasserstein Generative Adversarial Networks

Shouvanik Chakrabarti, Huang Yiming, Tongyang Li, Soheil Feizi, Xiaodi Wu

Abstract: The study of quantum generative models is well motivated, not only because of its importance in quantum machine learning and quantum chemistry but also because of the perspective of its implementation on near-term quantum machines. Inspired by classical adversarial training of generative models, we initiate the study of the Wasserstein Generative Adversarial Networks (WGANs) in the quantum setting. Specifically, we propose a definition of the Wasserstein semimetric between quantum data and formulate a concrete design of quantum WGANs. The entire implementation of our quantum WGAN is in principle efficient and hence scalable on quantum machines. We demonstrate that our quantum WGANs inherit a few key theoretical merits of classical WGANs. Our numerical study, via classical simulation of quantum systems, shows the more favorable numerical performance of our quantum WGANs over other quantum GAN proposals. We also discuss the possibility of the implementation of quantum WGANs on near-term machines by analyzing their performance with noisy quantum operations.

Code Link: https://github.com/yiminghwang/qWGAN

CMT Num: 3674