The homeworks should be done individually. You
should solve and write the problems up yourself. Discussing basic
techniques with classmates is OK, but
its best to never make
written notes while discussing individual homeworks.
Leading Class Discussions on Papers
You should be prepared to lead a discussion on the following general questions:
You can prepare a short summary or handout in advance, if you wish;
you needn't prepare a talk but you can do so if you find it useful.
You can consult other papers that might be interesting to use as
background or for contrast, if you wish, but this isn't required.
- What problem did the paper address?
- Is it important/interesting?
- What is the approach used to solve the problem?
- How does the paper support or otherwise justify the conclusions it reached?
You should aim for a total of 20 minutes, perhaps half your lead and
half class discussion. You should not assume that class members have
read the paper in detail.
You should also pick out relevant important sections of each paper
with regard to the above that we might want to take turns reading
aloud and discuss in class.
Taking Part in Class Discussion of Papers
Everyone should be reading the papers assigned for that day in
advance. You are not expected to understand the details of the papers,
but are expected to try to get a good overall sense of each paper.
While you are reading each paper, you should develop at least one
question or point relevant for discussion of the paper.
Writing Weekly Reviews of Papers
Each week, you will write a short review about one aspect of one paper
assigned that week. You needn't write a review the week you present
your paper. You should pick a topic that intrigued you or made you
think, and that can be discussed in class when we discuss the
paper. It should be at most 1 page in length. Possible topics are as
follows, but feel free to choose others.
- Was the paper, in your estimation, successful in what it attempted?
Why or Why not?
- What did you learn from the paper?
- Is there interesting future work, or are there new ideas suggested by the paper for solving this problem, or others?