Explore cutting-edge machine learning strategies for neural PDE solvers in this hour-long talk by UC Berkeley Assistant Professor Aditi Krishnapriyan. Delve into self-supervised learning techniques for solving fluid dynamics and transport PDE problems using spectral methods, and discover "simulation-in-the-loop" approaches that incorporate PDE-constrained optimization as a neural network layer. Learn how machine learning methods can be seamlessly integrated with numerical methods in fully differentiable settings, addressing challenges in spatiotemporal modeling and limited data scenarios. Gain insights into physics-inspired machine learning, geometric deep learning, and inverse problems, with applications in molecular dynamics and fluid mechanics.
Bridging Numerical Methods and Deep Learning with Physics-Constrained Differentiable Solvers
Inside Livermore Lab via YouTube
Overview
Syllabus
DDPS | Bridging numerical methods and deep learning with physics-constrained differentiable solvers
Taught by
Inside Livermore Lab
Related Courses
-
Neural Differentiable Physics: Unifying Numerical PDEs and Deep Learning for Data-Augmented Computational Physics
-
Deep Learning and Numerical Methods Intersections for Improving Molecular and Fluid Dynamics
-
Model-Constrained Deep Learning Approaches for Scientific Machine Learning
-
Deep Learning and Computations of PDEs by Siddhartha Mishra
-
PirateNets - Physics-Informed Deep Learning with Residual Adaptive Networks
-
Scientific Machine Learning through the Lens of Physics-Informed Neural Networks
