High Order Recovery of Geometric Interfaces from Cell-Average Data
Albert Cohen
Professional Summary
About Me
I’m very enthusiastic about science and the possibility of using artificial intelligence to improve society. I have a strong background in physics, networks, data science and modelling. I’m particularly interested open science and new ways of creating and sharing knowledge.
Education
PhD in applied mathematics
Sorbonne Université
Master in Physics
University of Buenos Aires
Bachelor in Physics (6.5 years program)
University of Buenos Aires
Interests
Scientific Machine Learning
Reduce Order Modelling
Optimization
Optimal sampling
📚 My Research
My research currently focuses on developing accelerated version of Natural Gradient Descent specially in the context of Scientific Machine Learning and Non-linear Reduced Order Modelling.
Featured Publications
•
•
1 min read
Publications
(2025).
High Order Recovery of Geometric Interfaces from Cell-Average Data.
ESAIM: Mathematical Modelling and Numerical Analysis.
(2024).
Model Reduction for Forward Simulation and Inverse Problems : Towards Non-Linear Approaches.
(2024).
State Estimation of Urban Air Pollution with Statistical, Physical, and Super-Learning Graph Models.
Advances in Computational Science and Engineering.
(2023).
Nonlinear Compressive Reduced Basis Approximation for PDE's.
Comptes Rendus. Mécanique.
Recent & Upcoming Talks
Talk at LMJL seminar of applied math
Natural gradient descent with momentum
Talk at MORTech 2025
Natural gradient descent with momentum for physics informed learning
Talk at minisimposium in ENUMATH 2025
Natural gradient descent with momentum