Deep Learning and Computations of PDEs by Siddhartha Mishra

Explore a distinguished lecture on deep learning applications in partial differential equations (PDEs) and high-dimensional computations. Delve into topics such as uncertainty quantification, supervised learning with deep neural networks, PDE constrained optimization, and operator learning. Gain insights on refined error estimates, training on low-discrepancy sequences, and bounds on reconstruction, encoding, and approximation errors. Examine real-world applications like tsunami modeling in the Mediterranean Sea and learn about challenges in out-of-distribution evaluations. Participate in a live interactive session with the speaker, where you can submit questions in advance through a provided Google form.

Intro
Partial Differential Equations (PDES)
Uncertainty Quantification
Supervised learning with Deep Neural networks
Supervised learning for high-d Parametric PDES
Refined Error Estimates
Training on Low-Discrepancy Sequences
PDE constrained Optimization
Tsunami in the Mediterranean sea
DL for Many-Query Problems: Further Issues
Operator Learning
DeepOnet Decomposition
Bounds on Reconstruction Error
Bounds on Encoding Error
Bounds on the Approximation Error
Out of Distribution Evaluations
Deep learning and High-dimensional PDES