Media Summary: Talk by Pascal Germain at NIPS 2012 Workshop Multi-trade-off in Benjamin Guedj (2021), A (condensed) primer on In this lecture we introduce a compression approach to obtain bounds for test-train risk difference. We prove a
Part 2 Pac Bayesian Learning For Deep Learning - Detailed Analysis & Overview
Talk by Pascal Germain at NIPS 2012 Workshop Multi-trade-off in Benjamin Guedj (2021), A (condensed) primer on In this lecture we introduce a compression approach to obtain bounds for test-train risk difference. We prove a Speakers: Andrew Foong, David Burt, Javier Antoran Abstract: Olivier Catoni - Dimension-free PAC-Bayesian Bounds (Talk)