Seth Lloyd: Quantum Generative Adversarial Networks

A talk by Seth Lloyd at the Quantum Machine Learning Workshop, hosted September 24-28, 2018 by the Joint Center for Quantum Information and Computer Science at the University of Maryland (QuICS).

Abstract: In generative adversarial learning, a generator attempts to generate `fake data' to fool a discriminator, who in turn tries to distinguish between the fake data and the real data. The learning process can be regarded as an adaptive game, which converges to a fixed point where the generator generates data with the same statistics as the real data. This talk investigates the quantum version of generative adversarial networks, where the data may be classical or quantum, and the generator and discriminator may have access to quantum information processing. I show that the fully quantum adversarial learning game converges to a point where the generator successfully mimics the quantum data, but that without full quantum capability, the generator will fail.