English version below
Bonjour,
Nous avons le plaisir de vous annoncer l’école d’automne “Recent advances in Machine Learning” du 16 au 21 novembre 2025 au centre Paul Langevin d’Aussois. Organisée à l’initiative du groupe “Statistique Bayésienne” de la SFdS, le programme s’articulera notamment autour de mini-cours sur des notions modernes en apprentissage statistique comme les modèles génératifs (Marylou Gabrié), l’apprentissage fédéré (Aymeric Dieuleveut), les attaques par exemple adverse en IA (Rafaël Pinot) et l’optimisation (Adrien Taylor).
L’école accueillera également une session poster et des présentations courtes par les participant·e·s. Nous encourageons vivement les doctorant·e·s et post-doctorant·e·s à soumettre un abstract.
Les inscriptions sont ouvertes à toutes et tous, vous trouverez le programme détaillé ainsi que la procédure sur le site web : https://mlataussois.sciencesconf.org
Voici un rappel des dates importantes :
Les frais d’inscription sont fixés à 575€ pour les non-permanent·e·s, et couvrent l’intégralité de la pension complète.
Cet événement est en partie sponsorisé par DATAIA, le département MathNum d’INRAE, le CNRS, et bénéficie du soutien logistique de la SFdS.
Au plaisir de vous accueillir à Aussois,
Le comité d’organisation : Pierre Gloaguen, Nicolas Jouvin, Julien Stoehr (président)
Le comité scientifique : Christophe Denis, Anna Korba, Sylvain Le Corff (président)
-------------------------------------------------------------------------------------------------------------------------------
Hello,
We are pleased to announce the “Recent advances in Machine Learning” autumn school from November 16 to 21, 2025 at the Paul Langevin center in Aussois. Organized on the initiative of the SFdS “Bayesian Statistics” group, the program will feature mini-courses on modern notions in statistical learning such as generative models (Marylou Gabrié), federated learning (Aymeric Dieuleveut), evasion attacks in AI (Rafaël Pinot) and optimisation (Adrien Taylor).
The school will also host a poster session and short presentations by participants. We strongly encourage PhDs and postdocs to submit an abstract.
Registration is open to all, and details of the program and procedure can be found on the website: https://mlataussois.sciencesconf.org
Here's a reminder of the key dates:
The registration fee is €575 for non-permanent students, and covers full board.
This event is partly sponsored by DATAIA, INRAE's MathNum department and CNRS, and benefits from logistical support from SFdS.
We look forward to welcoming you in Aussois,
The organizing committee: Pierre Gloaguen, Nicolas Jouvin, Julien Stoehr (chairman)
Scientific Committee: Christophe Denis, Anna Korba, Sylvain Le Corff (chairman)