Mini-colloque 9 – MC09
Applications de l'Apprentissage Automatique en Physique
Atomic Molecular Physics and Optics Division
Stéphane Barland (Université Côte d'Azur stephane.barland@univ-cotedazur.fr), Emmanuel Centeno (UCA emmanuel.centeno@uca.fr) et Demetrio Macias (UTT-CNRS demetrio.macias_guzman@utt.fr)
Artificial intelligence and physics are closely related in many ways. AI obviously opens new avenues for physics research, from model discovery to optimization or control of experimental systems. In turn, physics can contribute to the understanding of learning mechanisms or to the realization of computational tasks typically entrusted to AI algorithms.
To explore these links, three sessions will be organized,
1. Control, inverse problems and experimental physics
The aim of this session is to bring together experts to share their experiences in using AI to process experimental data, solve inverse problems and control dynamic systems.
Guest speaker: Gaël Varoquaux (Inria Saclay-Île de France, Palaiseau, France) - Title to be announced.
2. Model discovery and computation
In this session, we'll be looking at how the processing of massive data can be used to establish a relationship between the different variables of a given problem, and to gain a better understanding of it when models are unreliable or non-existent.
Guest speaker: Rodrigo Ibata (Strasbourg Astronomical Observatory, University of Strasbourg) - Title to be announced.
3. Physics of learning
In this session, we aim to bring together experts in the reciprocal links between physics and machine learning, focusing on computation on physical substrates, emerging phenomena in neural networks and, more generally, the physics of learning.
Guest speaker: Rémi Monasson (Laboratoire de Physique Théorique de l'ENS) - Title to be announced.
Wherever possible, the preferred language for oral presentations and posters will be English.
This mini-colloquium is supported by the GdR I-GAIA, Biocomp and Ondes.
