AI for Science Seminar (FS2024)
Artificial intelligence (AI) and machine learning (ML) offer significant potential to revolutionize the fundamentals of scientific computation and discovery today. The goal of this seminar course is to expose student to the recent development of "AI for Science".
Lecturers: Prof. Niao He (OAT Y21.1), Prof. Zebang Shen (OAT Y21.2).
Kick-off Meeting: February 21, 2024; 12:15-14:00; CAB G 56
Presentation Sessions:
May 25 8:30-13:30 (OAT Seminar Room),
June 1 8:30-13:30 (OAT Seminar Room).
Useful Links: Course Catalogue (252-5256-00L). All materials are available at Moodle including annoucements and Q&A.
Presentations
| Group ID | Presenter(s) | Paper Title | Presentation Slides |
|---|---|---|---|
| 1 | Joel Reimann Timon Wattenhofer |
Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations | file |
| 2 | Sonja Joost Gohar Tamrazyan |
B-PINNs: Bayesian physics-informed neural networks for forward and inverse PDE problems with noisy data. | file |
| 3 | Luca Apolloni Eren Homburg |
Fourier Neural Operator for Parametric Partial Differential Equations. | file |
| 4 | Simon Storf |
Choose a transformer: Fourier or galerkin. | file |
| 5 | Khaled Kerouch Jinwei Zhang |
The Deep Ritz Method: A Deep Learning-Based Numerical Algorithm for Solving Variational Problems. | file |
| 6 | Marcel Maciejczyk Franz Schwinn |
Solving high-dimensional partial differential equations using deep learning. | file |
| 7 | Yifei Han Jonathan Seele |
Understanding and mitigating gradient flow pathologies in physics-informed neural networks. | file |
| 8 | Nicola Witzig Andrin Zoller |
Self-consistent velocity matching of probability flows | file |
| 9 | Lino Hofstetter Damiano Meier |
A Tutorial on the Non-Asymptotic Theory of System Identification | file |
| 10 | Zhiang Chen Longxiang Jiao |
Hidden fluid mechanics: Learning velocity and pressure fields from flow visualizations. | file |
| 11 | Gaspard Krief Emilia Pucher |
Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators | file |