Recently, physics-informed learning (PIML) approaches have been highly successful in many fields ranging from physics to life and environmental sciences. These methodologies enable the hybridization of physical knowledge of the field and real data in order to solve a variety of problems. This national conference aims to bring together the various communities working on physics-informed ML approaches. The program will revolve on four main themes:

• statistical aspects
• inverse problems
• neural operators
• modeling & numerical analysis 

Where & when?

The conference will take place on Wednesday, April 15, 2026, on the AgroParisTech campus in Palaiseau, Amphi A0.04. See the “Getting to AgroParisTech” page for more details on how to get there.

Who can participate?

The event is open to everyone, subject to availability. Registration is free but mandatory before March 31, 2026. We strongly encourage participation from the various communities involved in PIML: machine learning, statistics, modeling/PDE, physics, etc.

The symposium will run from 9 a.m. to 5 p.m. + a poster session from 5 p.m. to 6 p.m. (here to submit a poster).

The speakers are:

•     Claire Boyer (Université Paris-Saclay) 
•     Clémentine Courtès (Université de Strasbourg)
•     Lucas Drumetz (IMT Atlantique)
•     Lise Le Boudec (Sorbonne Université)
•     Jordan Patracone (Télécom Saint-Etienne)
•     Arnaud Vadeboncoeur (University of Cambridge)

Click here for the detailed program for the day.

In addition, a satellite tutorial on PIML will take place the day before, Tuesday April 14th, from 2 to 5:30 p.m. at AgroParisTech.  

• Registration is free but mandatory. Click here to register
• A vegetarian lunch will be provided.
• How to get to AgroParisTech? Visit this page

The event benefits from financial and logistical support from AgroParisTech, the DATAIA Institute, the Fondation Mathématique Jacques Hadamard (FMJH), and the Industrial Data Analytics and Machine Learning (IDAML) Chair.

© Affiche - Daniele Clarotto