Topics will vary from semester to semester. See Computer Science Division announcements.

Machine learning in the context of scientific problems is an exciting emerging area of research, and often requires the development of new methods that can incorporate and exploit the inductive biases and structure needed for such problems. There are also now numerous examples of concepts in physics historically influencing machine learning methods development more broadly. This course will give an overview of different physics-inspired machine learning methods, and the connections between concepts in physics (numerical methods, dynamical systems, symmetries, conservation laws) and machine learning. Topics include: overview of different types of physical constraints for neural networks, connections between dynamical systems and neural networks, implicit differentiation and differentiable physics, geometric deep learning including graph neural networks and incorporating symmetries into neural networks

Requisites

• Students not in the Master of Engineering Program

Repeat Rules