Course Websites

Subject offerings of new and developing areas of knowledge in theoretical and applied mechanics intended to augment the existing curriculum. See Class Schedule or departmental course information for topics and prerequisites. Course Information: May be repeated in the same or separate terms if topics vary to a maximum of 12 hours.

Through readings and projects, we will explore how the foundations of group representation theory can be used to construct symmetry-preserving algorithms for use in machine learning applications. Topics include: group theory, Euclidean and permutation groups, group representations (reducible and irreducible), regular group convolutional neural networks, steerable group convolutional neural networks, and equivariant graph neural networks. There will be an emphasis on application to the physical sciences, including problems in materials. Projects will focus on building algorithms for graphs, scientific data, and other data to preserve desired symmetries. Prerequisites: 1) Math 257 Linear Algebra with Computational Applications (or equivalent) (2) CS 125 Intro to Computer Science (or equivalent) (3) CS 444 Deep Learning for Computer Vision (or similar, introductory experience with convolutional and/or graph neural networks desirable)