Skip to main content

Course project

Students will complete a semester-long research project related to machine learning-augmented algorithm design. Projects should involve both theoretical and empirical results - e.g., developing a new algorithm or improving on an existing algorithm for some problem and proving theoretical guarantees on its performance, as well as evaluating it experimentally.

Project format

Projects will be conducted in groups of 1-3; you should discuss with your classmates early on to find topics of common interest and form project groups. Please attend office hours or schedule a meeting with the instructor if you need project topic suggestions. The ideal project will yield a short paper of quality comparable to a workshop paper at, e.g., NeurIPS; projects should be scoped accordingly.

Project deliverables include:

Milestones and Timeline

Milestone Due date Details Fraction of project grade
Project proposal October 8 Groups are highly encouraged to meet with the instructor before the proposal deadline to ensure the topic is well-scoped and relevant to the course topic 1/6
Mid-semester project update presentation November 5 Overview of topic, progress, and next steps 1/6
Final presentation December 15, 9am-12pm Presentation of project and results 1/3
Final report December 15, end of day Complete project write-up 1/3

Grading rubric

Check back soon for a more detailed grading rubric for each of the project deliverables

Project resources