Geometry, topology and machine learning

The conference will serve as a gathering point for researchers interested in studying machine learning and deep learning from a geometric perspective and, more generally, a mathematical viewpoint, as well as applying geometric and mathematical tools in machine learning in a broad sense. Geometric is understood in a broad sense, encapsulating but not limited to algebraic and topological tools for continuous and discrete structures. 

This workshop follows a Geometry and Machine Learning held in Scalea, Italy, in 2024 and of Geometry and Machine Learning 2 held in Paris, France in 2025.

The conference will be held in Paris at Sorbonne Université, Paris, from May 18 to May 22, 2026.

To give a talk please fill in the following form : link to the form.

For any questions, please contact: gregoire.sergeant-perthuis@sorbonne-universite.fr 

Practical informations: 

Where: Room C404, CQSB, Building C, 4th floor.
Sorbonne Université
4 Place Jussieu, 75005 Paris

Metro Station: Line 10, or Line 7 metro station Jussieu.

To schedule your itinerary a useful link is here.

Confirmed speakers

Tolga Birdal  (Imperial College London)
Levent Doğan (Faculty of Computer Science, Ruhr University Bochum)
Argyris Kalogeratos (Université Paris-Saclay, ENS Paris-Saclay, CNRS)
Jiayi Li (MPI-CBG)
Steve Oudot (Inria Saclay)
Josué Tonelli Cuento (Johns Hopkins University)