AI is having a profound impact on science by accelerating discoveries across physics, chemistry, biology, and engineering. This course aims to present a highly topical selection of state-of-the-art AI applications across these fields. Emphasis will be placed on using AI, particularly deep learning, to understand physical and engineering systems, mathematically modelled by PDEs.

Topics include physics modeled by PDEs and the limitations of traditional simulators; neural PDE solvers (PINNs and variants); neural operators (FNO, CNO, Operator Transformers); graph neural networks and flexible Transformer frameworks for complex geometries; generative AI (diffusion, flow) for multiscale problems and uncertainty quantification; physics foundation models; downstream applications such as uncertainty quantification, inverse problems, design, and AI for weather and climate; and selected examples in chemistry and biology (GNNs and generative AI for structure-based drug design). By the end of the course, students will be familiar with advanced applications, core algorithmic design and theory, the trade-offs of using AI for scientific and engineering problems, and key scientific machine learning themes.

  • Lectures: Thu 08:15–10:00 • ML H 44
  • Tutorials: Mon 12:15–14:00 • ML H 44
  • Discussion and submissions: Moodle