MS191A Exploring New Avenues for the Interaction of Numerical Methods for PDEs and Deep Learning I
MS Corresponding Organizer:
Mr.
Paolo Zunino
(
MOX, Department of Mathematics, Politecnico di Milano
, Italy
)
Chaired by:
Mr. Paolo Zunino (MOX, Department of Mathematics, Politecnico di Milano , Italy) , Prof. Kent-Andre Mardal ( Simula Research Laboratory / University of Oslo , Norway)
Mr. Paolo Zunino (MOX, Department of Mathematics, Politecnico di Milano , Italy) , Prof. Kent-Andre Mardal ( Simula Research Laboratory / University of Oslo , Norway)
Scheduled presentations:
-
Data driven approximation of the solutions to parametric eigenvalue problems
-
DeepONet for Matrix-Free Preconditioners Construction
-
Learning Mesh Extension Operators
-
Collocation method with two-layer Neural Networks
-
Deep-Learning Preconditioners for Mixed-Dimensional Partial Differential Equations on 3D-1D Domains
-
Enhancing Thermal Conductivity Modeling of Polyurethane with Phase Change Materials via Physics-Informed Neural Networks at Multiple Scales
