Colloque - Andrea Manzoni : Reduced Order Modeling and Scientific Machine Learning: Synergies and Opportunities

Épisode de Informatique et sciences numériques (2025-2026) - Yvon Maday par Collège de France · 23 juin 2026 · 52 min

Yvon Maday

Chaire Informatique et sciences numériques

Collège de France

Année 2025-2026

Colloque : Aspects mathématiques et appliqués des méthodes de réduction de complexité - Andrea Manzoni : Reduced Order Modeling and Scientific Machine Learning: Synergies and Opportunities

Andrea Manzoni

Associate Professor of Numerical Analysis, MOX - Department of Mathematics, Politecnico di Milano, Italy

Résumé

Among several recently proposed data-driven Reduced Order Models (ROMs), deep learning-based ROMs (DL-ROMs) have proved to be a successful strategy to construct non-intrusive, highly accurate surrogates for the real time solution of parametric nonlinear time-dependent PDEs. By relying on (possibly, convolutional) autoencoders, it is indeed possible to generate latent spaces where the candidate solution is then sought, as a function of parameters and time, using an additional neural network.

In this talk I will provide an overview on DL-ROMs, discussing some recent theoretical results that justify their construction, and connecting them to classical reduced basis methods. Then, I will showcase a series of possible extensions of DL-ROMs capable to (i) handle knowledge of physical laws, (ii) deal with varying geometries, (iii) identify the latent dynamics to ensure accurate out-of-training forecasts, and (iv) include uncertainty quantification.

In all these cases, we will show how the construction of a suitably expressive—and possibly explainable—latent space is essential to ensure accuracy and efficiency of reduced order models exploiting deep neural networks, drawing also some conclusions of possible interest to other contexts in scientific machine learning.

Autres épisodes de « Informatique et sciences numériques (2025-2026) - Yvon Maday »

À propos du podcast

Créée en partenariat avec Inria, la chaire annuelle Informatique et sciences numériques marque une volonté commune de faire valoir l'importance de cette discipline scientifique et la nécessité de lui octroyer une place pleine et entière. Réduction de complexité pour les simulations numériques…

Informatique et sciences numériques (2025-2026) - Yvon Maday

Épisodes similaires

Tous les épisodes de Informatique et sciences numériques (2025-2026) - Yvon Maday · Plus de podcasts Éducation