Danial Faghihi, Serge Prudhomme, and I are preparing a proposal for a short course, Neural Operators: Foundations and Extensions, for USNCCM19. I would like to get input from the community before we finalize the topics.
We plan to cover the mathematical background and implementation of different neural operator architectures, followed by Bayesian neural networks for uncertainty quantification and multi-level neural networks for improving accuracy. We are also planning a hands-on session using Claude to develop and check an example. We would share a Claude Artifact afterward and provide all lecture materials in a dedicated GitHub repository.
I would welcome suggestions from both experts and those interested in learning the topic, especially anyone who has wanted to try neural operators.
- What should we emphasize in the mathematical foundations and implementation? Are there particular concepts or methods you would like us to work through in detail, with examples and code?
- Which applications should we highlight where neural operators have been particularly effective? Examples from your own work, including what made them useful for that problem, would be especially helpful.
- What limitations have you encountered that deserve more attention? I would like us to discuss these and the research questions they raise.
I would love to hear your thoughts!
| Attachment | Size |
|---|---|
| Flyer | 1.17 MB |