Academics

  /  

Courses

  /  

Descriptions

COMP_SCI 497: Recent Highlights in the Theory of Machine Learning

Description

This is a graduate paper reading course, where every week we'll have the presenters covering 1-2 papers on a related topic. The aim of the course will be to cover interesting, influential papers in learning theory (with a particular emphasis on deep learning theory) . There are two goals (1) explore different frameworks, theories, and problems towards getting a better understanding of modern machine learning, and (2) getting a flavor of some of the mathematical techniques that are useful to do research in this area. Hence the classes will consist of blackboard talks with proof details to the extent that this is possible in a couple of hours, and will hopefully stimulate further discussions on open problems related to the paper. Details of the topics and papers will be posted soon on Canvas. This is also related to the IDEAL Spring Special Program on Theory of Deep Learning and Optimization.

Class logistics: Tuesday 12:30pm - 3:20pm in Tech F280. First class: April 8, 2025 (since April 1 follows Northwestern Monday).

Pre-Requisites: CS PhD students or permission of instructor. The student is expected to have already taken a graduate course in theory, like Graduate Algorithms.

• This course fulfills the Project or Technical Elective area.

REFERENCE TEXTBOOKS: N/A
REQUIRED TEXTBOOK: N/A

COURSE COORDINATORS: Aravindan Vijayaraghavan

COURSE INSTRUCTOR: Aravindan Vijayaraghavan