Course Code at McMaster University: MATH 4AI3-C01

Registration Deadline: September 20, 2026

Instructor: Professor Anastasis Kratsios, McMaster University

Course Dates: September 9 - December 10, 2026
Mid-Semester Break: October 12-16, 2026
Lecture Times: Mondays, Wednesdays, & Thursdays | 10:30 AM - 11:20 AM (ET)
Office Hours: Wednesdays | 5:00 PM - 6:00 PM (ET)

Registration Fee:

• Students from our Principal Sponsoring & Affiliate Universities: Free
• Other Students: CAD$500

Capacity Limit: 60 students (auditing is allowed for this course; please indicate in your registration if you are auditing)

Format:

• In-Person: McMaster University
• Online: via Zoom

Course Description

A rigorous introduction to the mathematical foundations of modern artificial intelligence and deep learning, with an emphasis on approximation theory, learning theory, and high-dimensional statistical methods. 

This course develops the mathematical foundations needed to understand modern deep learning from a rigorous theoretical perspective. Particular emphasis is placed on algorithmic approximation theory and empirical-process methods for learning in high dimensions. Topics include approximation and expressivity of neural networks, probabilistic and statistical tools for learning, generalization, optimization, and selected mathematical aspects of modern deep-learning architectures.

Prerequisites: A strong undergraduate background in analysis, linear algebra, and probability; mathematical maturity appropriate for a fourth-year mathematics course.

Evaluation Method: Final projects and mid-term competition.

Participation Grades for joining Fields AI Seminar: http://www.fields.utoronto.ca/activities/26-27/mathai27

Instructor Website: https://anastasiskratsios.github.io/

Course Notes/Proto-Book: https://drive.google.com/file/d/18DZ4KJOXIez66_1h3jziYw8COlDrHgj2/view?u...