MATH 581: High Dimensional Probability and Statistical Learning

Instructor Information

Instructor: Dmitriy Drusvyatskiy
Lecture: MW 2:30-3:50 PM
Office hours: W 1:00-2:00 PM and by appointment
Office: PDL C-434
Email: ddrusv at

Meeting Times and Location

lecture time: Monday and Wednesday 2:30-3:50 PM
lecture location: Raitt Hall (RAI) 116

Course Description

This is an introductory course in high-dimensional probability and statistical learning. The main focus will be on the concentration phenomenon in high dimensions and its consequences for inference and classification from limited data. In particular, we will use the developed techniques to analyze algorithms for various statistical inverse problems, such as community detection, sparse recovery, low-rank matrix completion, phase retrieval, etc. This course is appropriate for anyone with a working knowledge of linear algebra, probability, and mathematical analysis.


We will primarily use elements from the following two texts:

Requirements and Grading


You may work together on problem sets, but you must write up your own solutions. You must also cite any resources which helped you obtain your solutions.

Weakly Schedule/Notes (evolving)

The following is a very rough schedule for the first 6-7 weeks of the course. We will decide on what to cover in the last three weeks, depending on the students’ interests.

Homework Problems