Title: A primer on PAC-Bayesian learning, and some application
Speaker: Benjamin Guedj (https://bguedj.github.io/)
Abstract: PAC-Bayes is a generic and flexible framework to address generalisation abilities of machine learning algorithms. It leverages the power of Bayesian inference and allows to derive new learning strategies. I will briefly present the key concepts of PAC-Bayes and illustrate a few of its recent successes (including generalisation guarantees for deep neural networks).
