MATH 7740

Statistical Learning Theory

Course Description

Course information provided by the 2025-2026 Catalog.

Learning theory has become an important topic in modern statistics. This course gives an overview of various topics in classification, starting with Stone's (1977) stunning result that there are classifiers that are universally consistent. Other topics include plug-in methods (k-nearest neighbors), reject option, empirical risk minimization, Vapnik-Chervonenkis theory, fast rates via Mammen and Tsybakov's margin condition, convex majorizing loss functions, RKHS methods, support vector machines. Further, active high-dimensional statistical research topics such as lasso type estimators, low-rank multivariate response regression, topic models, latent factor models, and interpolation methods are presented.

Prerequisites/Corequisites Prerequisite: basic mathematical statistics (STSCI 6730/MATH 6730 or equivalent) and measure theoretic probability (MATH 6710).

Permission Note Enrollment limited to: graduate students.

Last 4 Terms Offered 2025FA, 2024FA, 2023FA, 2022FA

When Offered Fall.

View Enrollment Information