SGAP 2026 is a thematic school designed for researchers and PhD students in the Île-de-France region who want to get started with machine learning in their scientific practice. Over three days, participants will be introduced to the fundamentals of modern AI methods, both conceptually and practically, and hear from scientists who have successfully integrated these tools into their research.

The school targets scientists from all disciplines with little or no prior background in machine learning. It is intentionally pedagogical: speakers are chosen for their ability to make complex ideas accessible to a broad scientific audience.

Registration is free but mandatory. SGAP will host approximately 100 participants. Pre-registrations are open via the website, and full registration will require a confirmation step in October to best accomodate for the number of participants.

 

Organization The event is organized by Marylou Gabrié (ENS Paris), Thomas Moreau (Inria MIND), Davide Carbone (ENS Paris) and Mansour Benbakoura (INRIA MIND) with support from the DIM AI4IDF and of a scientific committee composed of Florence d’Alché-Buc (Télécom Paris), Julie Deshaye (Institut Pierre Simon Laplace) and Bertrand Thirion (INRIASaclay).

The first two days are dedicated to hands-on tutorials covering the foundations of machine learning and deep learning.

Day 1 -- Statistical Learning

• Morning: Introduction to the core concepts of statistical learning — Gaël Varoquaux (Inria)
• Afternoon: Practical introduction to machine learning in Python with scikit-learn — Guillaume Lemaitre (Inria)

Day 2 -- Deep Learning

• Morning: Introduction to neural networks and deep learning — Vicky Kalogeiton (École Polytechnique)
• Afternoon: Practical session on deep learning with scientific data — Thomas Moreau (Inria) & Ernest Mordret (Institut Pasteur)

The third day showcases concrete, successful examples of AI applications across scientific disciplines. Each talk is designed to be accessible to non-specialists and to spark ideas across fields.

Confirmed and anticipated speakers:

• AI & climate — Anastase Charantonis (Inria)
• AI & genomics — Flora Jay (Université Paris-Saclay)
• AI & astrophysics — Erwan Allys (ENS Paris)
• AI & medical imaging — Émilie Chouzenoux (Inria Saclay)
• AI & quantum physics — Filippo Vicentini (École Polytechnique)
• AI & softmatter — Sophie Marbach (Mistral AI)
• AI & chemistry — to be confirmed