I've set up a google
group where you can get latest information on the current teaching
activities, ask questions, and discuss issues. You have to become a
member to view the group contents, but access is not restricted.
During the winter term 2007/2008 I was on parental leave.
This script has been written by Paul Bünau and me as
supplementary material to the lab course on machine learning. In
it we try to follow a "hands-down" approach to machine
learning. For each method, you find a categorization, a brief
explanation of how the method works, and, most importantly, pseudo
code to implement the method quickly.
This talk gives introduces the basic ideas behind Bayesian
inference (that is, Bayes rule). The concept of conjugate prior is
discussed on some simple distributions, including how to "guess"
the conjugate prior to some distribution. Finally, some more
high-level discussion of the differences between Bayesian and
frequentist approaches are discussed.
Disclaimer: I'm not a Bayesian
This talk gives a short overview over machine learning, including
typical applications (well, mostly applications from our group).
This is actually something I've been thinking about for some
time. I always thought that having only a "naive" understanding of
probability theory becomes a bit limiting at some point and that
it would be helpful if you had seen at least once how probability
theory is "built" in mathematics. Therefore, in this talk I try to
introduce all the normal concepts but also give some idea of
what that means mathematically.