Assessment in this class is based on one presentation (worth 50 percent of your grade) and one project (worth another 50 percent).

The presentation is an individual 20-min presentation based on a paper of your choice. For ideas for selecting papers, see the Reading List tab. You are also very welcome to select papers not on the reading list; please run them by me first. Paper selections and presentation date preferences are due by Apr 24. You are very welcome to select a paper that is related to your project. You should send me your presentation slides for feedback at least one week before the presentation. If for some reason you cannot make an in-class presentation, please send me an email.

The project is a mini-research project, and is to be done in groups of size one or two. A project report is due at the end of the quarter.

The project can be of a few different types. You can pick a paper and apply the method in the paper on one or more new datasets (that were not tried in the paper), and draw some observations. Most papers have code, so this should be doable. You can also compare methods proposed by a few different papers on some new datasets, and draw some conclusions. You can try to prove something interesting about a method proposed by a paper of your interest. Or you can try to establish some new conceptual connections between some of the topics we discuss in this class.

The Friday lectures will be project meetings, and you are very welcome to come talk to me about project ideas - I'll be happy to make suggestions. Project teams are due by Apr 24, and topics by May 7.