The main expected outcome of the course is a good research project that makes a reasonable contribution to the field. Some possibilities include (but not limited to):

  • Developing a new algorithm that addresses a specific weakness of existing approaches (and getting near SOTA).
  • Exploring RL applications in domains like robotics, vision, NLP, smart grids, traffic, finance, biology etc.
  • Contributing highly stable implimentations of recent algorithms that reproduce results from the papers.

It is very important to discuss project ideas from an early stage with the instructors and start making progress. The best time to discuss this would be office hours, but appointments can also be made over email. All reports are expected in the NIPS format.

Apr 18th Project proposal (2 pages)
May 7th Milestone (4 pages)
Jun 6th Final report (8 pages max)
Jun 5th Project poster presentations

Posters should be 40 by 32 inches and can be landscape or portrait.