OBJECTIVE: To understand the various optimization techniques underlying the machine learning methods that have become so popular in real-world applications today. This course will provide an introduction to these methods, a discussion of their use, and opportunities to improve them as part of course projects. (This will be an interaction-oriented course devoted to a deeper understanding of optimization in machine learning. Expect the course to have mathematical rigor.)
THINGS: Introduction to optimization, Convex sets, convex functions, Lagrange duality, convex optimization algorithms, second order cone models, semidefinite programming, semidefinite programming, minimax, sublinear algorithms, inner point methods, active set, stochastic gradient, Coordinate descent, Cutting planes method, Image / video / multimedia processing applications
ELIGIBILITY: Students who have completed a basic machine learning course and are interested in learning its mathematical foundations. Prior instructor approval is required to register for the course.