Advanced topics in the theory of machine learning


Spring 2021

Harvard CS 229br: Boaz Barak Mondays 12-3pm Eastern (First lecture: Monday, January 25)

MIT 18.408: Ankur Moitra Wednesdays 12pm-3pm Eastern (First lecture: Wednesday, February 17)

Apply to one or both courses here. For full consideration, please apply by Wednesday January 21.

The two courses will cover modern topics in the theory of machine learning, and deep learning in particular. Both courses will contain both theorems and experimental results, but MIT 18.408 will emphasize mathematical foundations while Harvard CS 229br will focus more on experimental insights. We recommend students take both courses, which may share some lectures and assignments, but this is optional and students can choose to take either one or both of the courses.

Introductory blog post by Boaz: Machine Learning Theory with Bad Drawings

Students interested in these courses will likely also be interested in following our machine learning theory seminar series. You can sign up for the mailing list to get announcements and Zoom links. Also, since last spring, all the talks in this series have been recorded and available on the webpage.

Prerequisites (for both CS 229br and MIT 18.408): Both courses will require mathematical maturity, and proficiency with proofs, probability, and information theory, as well as the basics of machine learning. We expect that students will have both theory background (at Harvard: CS 121 and CS 124 or similar, at MIT: 6.046 or similar) as well as machine learning background (at Harvard: CS 181 or 183 or similar, at MIT: 6.036 or similar).

Apply for one or both courses: Both courses are open to Harvard and MIT graduate and undergraduate students. Both courses will have a limited number of slots. You can apply to both the Harvard and MIT courses by filling out this form. You can apply to one or both of the courses.