Winter 2026 STAT 37779

Course: STAT 37779

Title: Kernel Methods for Statistics and Machine Learning

Instructor: Anirban Chatterjee

Class Schedule: Sec 1: TR 3:30 PM–4:50 PM in TBA

Description: This course provides a rigorous introduction to kernel methods across statistics and machine learning, covering foundational concepts in kernels and reproducing kernel Hilbert spaces (RKHS), classical statistical tools such as kernel density estimation and kernel regression, and modern kernel-based inference techniques including discrepancy measures and hypothesis testing with MMD and related methods. The course then connects these ideas to contemporary machine learning by exploring the Neural Tangent Kernel (NTK) and kernel-based objectives in generative modeling, such as MMD GANs.