Algorithmic Statistics (6.S896), Fall ’23, MIT

Update: Course website moved!

new website

Old Course Website Below

Prerequisites: Mathematical maturity is the main prerequisite. Familiarity with linear algebra, probability, discrete math, and algorithms at the advanced undergraduate level will be assumed.

Meeting time: Tuesdays and Thursdays, 2:30pm-4:00pm

Location: 32-124

Instructors: Costis Daskalakis and Sam Hopkins

Office Hours: By appointment.

Evaluation: Students will be expected to complete two problem sets and a research-oriented course project, which may consist of original research (theoretical and/or experimental!) and/or an exposition of 1 or 2 recent research papers. Tentatively, weight for your final grade will be split as follows: 25% pset 1, 25% pset 2, 50% course project.

Lectures + Lecture Notes

No. Date Topics Notes/References
1 Sept. 7 introduction, uniformity testing on the hypercube notes part 1, notes part 2
2 Sept. 12 learning high-dimensional gaussians notes, O’Donnell lecture on Gaussians
3 Sept. 14 undirected graphical models notes


Homework problems