Séminaire départemental – On generalisation and learning – some theoretical results for deep learning – Vendredi 20 septembre 2024 – Benjamin Guedj

Vendredi 20 septembre 2024

On generalisation and learning – some theoretical results for deep learning
Benjamin Guedj
Principal Research Fellow in machine learning at University College London

Heure: 13h30
Local: PLT 2765

Résumé: Generalisation is one of the essential problems in machine learning and foundational AI. The PAC-Bayes theory has emerged in the past two decades as a generic and flexible framework to study and enforce 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 highlight a few recent contributions from my group focusing on the theory of deep learning.