Next week: I'll talk about the Long Path Problem

[R. Paul Wiegand]

Greetings,

Just to remind everyone of our tentative schedule. Next Friday, I'll
talk about something called the "Long Path Problem", a problem
contrived to demonstrate some properties in certain evolutionary
algorithms (e.g., genetic algorithms). Then we'll have a reading-
oriented discussion the following week on Markov Chains (see the other
post for that).

That doesn't get us very far into the semester, so I need volunteers
to step up, as soon as possible!

As far as next week goes: We'll start with the context for why the
Long Path Problem was invented, the empirical results obtained, then
go through some relatively straightforward analysis that makes the it
quite clear why it does not show what it was designed to show. The
moral of the story is: A little, simple analysis is sometimes
relatively easy, and can often give you much clearer answers than
empirical experiments.

R. Paul Wiegand.

[Guillermo]

If no one else has something to present the week after markov chains I
can talk about a randomized min-cut algorithm.

[R. Paul Wiegand]

Sounds good. We'll hear from you on randomized min-cut on Friday, Feb
4.