MS063B Neural Network Solvers for PDEs: Bridging Theory and Practice in Scientific Computing II
Main Organizer:
Dr.
Jooyoung Hahn
(
Czech Technical University in Prague / Faculty of Nuclear Sciences and Physical Engineering
, Czechia
)
Chaired by:
Dr. Alexander Heinlein (TU Delft , Netherlands) , Dr. Jooyoung Hahn (Czech Technical University in Prague / Faculty of Nuclear Sciences and Physical Engineering , Czechia)
Dr. Alexander Heinlein (TU Delft , Netherlands) , Dr. Jooyoung Hahn (Czech Technical University in Prague / Faculty of Nuclear Sciences and Physical Engineering , Czechia)
Scheduled presentations:
-
Keynote
Randomized Neural Networks for Solving PDEs
-
Frozen-PINNs: Can We Make Physics-Informed Neural Networks Fast, Accurate, and Causal?
-
Adding Constraints to Neural Networks: Applications to the Solution of PDEs
-
Accurate Physics-Informed Neural Networks for Fluid Dynamics: an Approximate Gram-Gauss-Newton Approach
-
Domain-decomposition inspired parallel training algorithms for scientific machine-learning
