Artem Maevskiy (HSE)

Generative Adversarial Networks and their application for fast simulation in HEP

Abstract: This talk will cover the basics of Generative Adversarial Networks (GANs), along with a few selected recent advancements. GANs allow for learning to sample from distributions without knowing their explicit forms, given just a set of example objects. An application of great interest for experimental high energy physics will also be discussed: using GANs for learning distributions from Monte-Carlo event simulation pipelines allows for drastic speedups of simulation processes.