AI in the Sciences and Engineering (2024) (camlab.ethz.ch)

AI in the Sciences and Engineering (2024) is a focused course from the ETH AI Center that surveys how deep learning is reshaping physics, chemistry, biology and engineering. It emphasizes scientific machine learning methods for systems governed by partial differential equations (PDEs), covering both practical implementations and the underlying theory. Students will learn current toolkits—surrogate and reduced-order models, physics-informed neural networks (PINNs), neural operators and operator learning—as well as how to design, implement and evaluate these algorithms in real scientific workflows. The course is significant because it bridges machine learning and traditional computational science, showing how data-driven models can accelerate discovery, enable fast multi-query simulation, and augment or replace costly numerical solvers. It also addresses critical challenges for deployment in science: uncertainty quantification, enforcing physical constraints, interpretability, data efficiency and generalization beyond training regimes. Taught at ETH Zurich with guest lecturers and applied-math teaching assistants, the program targets researchers who need both conceptual understanding and practical skills to responsibly apply AI to complex, PDE-modeled systems.