Postdoc
November 1, 2025
I am thrilled to announce that I started a postdoc at the Bocconi Institute for Data Science and Analytics (BIDSA).
I am currently a postdoctoral fellow at the Bocconi Institute for Data Science and Analytics (BIDSA), mentored by Giacomo Zanella. This position is funded by the ERC project "Provable Scalability for high-dimensional Bayesian Learning".
Before, I did my Ph.D. at the Department of Mathematics and Applications of ENS Ulm in Paris, working under the supervision of Gabriel Peyré and François-Xavier Vialard. This work was part of the Noria and Wolf ERC projects.
My thesis manuscript is available here.
Research Interests
My research field is theoretical Machine Learning: I study the mathematical properties of large-scale algorithms for solving problems in optimization, sampling and learning.
In my Ph.D., I used tools from the theory of optimization, optimal control and optimal transport to study the convergence of gradient based algorithm in the training of overparameterized neural network architectures such as deep ResNets.
Posts and news
Ph.D. Defense
October 10, 2025
I succesfully defended my Ph.D. whose thesis is entitled: “Training deep neural networks: mathematical models and convergence results”.
New preprint
July 23, 2025
We published a new preprint on “Ultra-fast feature learning for the training of two-layer neural networks in the two-timescale regime”.
Paper accepted at CPAM
July 23, 2025
I am very proud to announce that our paper on “Understanding the training of infinitely deep and wide ResNets with Conditional Optimal Transport” has been accepted for publication in Communications on Pure and Applied Mathematics.
Neurips 2022
November 28, 2022
I’ll be attending in person to the 36th Neurips conference in New Orleans, and will have the opportunity to present my work about “Global convergence of ResNets”.