Title:Adversarial training of quantum Born machine

Abstract:Adversarial learning is an effective way to train generative models, and has drawn lots of attention in the machine learning community in recent years. In the meantime, there are some researchs that use variational quantum circuits to generate simple patterns. In this paper, we study how to train a parameterized quantum circuit to generate the pattern distribution of a given training dataset, using the idea of generative adversarial network (GAN). Although our circuit architecture uses simple one-qubit rotation and two-qubit controlled-phase gates, the results of a small-scale proof-of-principle numerical experiment demonstrate that the quantum circuit can be trained in an adversarial way to generate the target distribution.