The project consists of reading a paper in the area of reinforcement learning, and then making a 30-45 minute presentation of the paper to the class.
The schedule for doing this will be posted soon, as will a suggested list of papers.