IML Machine Learning Working Group: special seminar on GANs

Friday 27 Oct 2017, 14:00 → 16:10 Europe/Zurich

222/R-001 (CERN)

Generative adversarial networks (GANs) are a recently introduced class of generative models, designed to produce realistic samples. They are relevant for High-Energy Physics to efficiently simulate events and detector interactions. First studies on calorimeter simulations promise improved simulation times by orders of magnitude. Ian will present current research on improving GANs in general, outside the HEP community.

Ian Goodfellow is Research Scientist working in machine learning at Google Brain, and is best known for inventing generative adversarial networks and for writing the textbook Deep Learning.

There will be a coffee break before the seminar, starting at 14:00. Following the seminar, there will be time for a general Q&A period, where students in particular are encouraged to ask questions.

14:00 → 14:30 Coffee break

14:30 → 15:15 Seminar: Open challenges for improving Generative Adversarial Networks (GANs)

Speaker: Ian Goodfellow (Google Brain)

2017-10-27-cern.pdf

Recording

15:15 → 15:25 Seminar questions

15:25 → 15:35 Break

15:35 → 16:00 General Q&A