STAT 6390

High Dimensional Probability and Statistics

Term: Fall 2026
Instructor: Yun Wei

Course description

Modern statistical applications often involve data sets in which the number of variables is comparable to, or substantially larger than, the number of observations. This course develops the mathematical and statistical foundations needed to analyze high-dimensional data.

Topics include concentration inequalities, random vectors and random matrices, high-dimensional geometry, sparse linear models, regularization methods, the Lasso, model selection, low-rank matrix estimation, principal component analysis, graphical models, minimax lower bounds, and computational considerations.

Particular emphasis is placed on how structural assumptions, including sparsity, low rank, and graphical structure, make statistically meaningful estimation possible in high dimensions. Students will use nonasymptotic techniques to derive finite-sample guarantees and evaluate estimators by consistency, prediction error, estimation error, and variable-selection performance.

Textbooks

  • Martin J. Wainwright, High-Dimensional Statistics: A Non-Asymptotic Viewpoint.
  • Roman Vershynin, High-Dimensional Probability (optional).

Lecture notes

The lecture notes evolve during the semester.

Download lecture notes (PDF)

Schedule

Date Topic
August 24 SubGaussian (Sec 2.1.1 and Sec 2.1.2) Scribed lecture notes
August 26 SubGaussian (Sec 2.1.2) and Sub-exponential (Sec 2.1.3) Scribed lecture notes
August 31 Symmetrization (Vershynin’s book Sec 6.3) and Sub-exponential (Sec 2.1.3) Scribed lecture notes
September 2 Sub-exponential, Bernstein Bounds, Jonson-Lindenstrauss Lemma (Sec 2.1.3) Scribed lecture notes
September 9 Sub-exponential (Sec 2.1.3) and Azuma-Hoeffding inequality (Sec 2.2.2) Scribed lecture notes
September 14 Bounded difference inequality (Sec 2.2.2) and Liptchize function of Gaussian variables
September 16 Liptchize function of Gaussian variables (Sec 2.3) and Entropic methods (Sec 3.1) Scribed lecture notes