finish a
programming project (to be announced later) (50% of the grade) write a term
paper at the end of the course (50% of the grade) ''Graphical Models'' is a course on graphical probabilistic
models. Graphical models are used for the structural aspects of a
problem (including dependency, cause and relevance) and techniques will
be developed for their analysis and interpretation. Examples will be
drawn from learning, information retrieval, location sensing, as well
as several other areas. The course starts with a brief
introduction to non-probabilistic graphical models and probability.
After that graphical probability models, independence modeling, (mostly
exact) inference in probabilistic graphical models, learning, and
applications are discussed. Both theory and applications are
covered.
Anyone interested in probabilistic modeling and dependency analysis for complex computational problems. The course is intended to give you new ideas, approaches or tools for your problems, and is aimed at senior undergraduates and graduate students. The course can be included in the M.Sc. or Ph.D. studies in the field of Adaptive and Intelligent Systems.
Note that the maximum number of participants is 25. You can registrate by filling in the registration form.
The course is an advances course and basic knowledge on probability (e.g. Probability I at the Math Department or Three Concepts: Probability) and basic graph theory (basic definitions such as directed graphs, arcs, paths, cycles, introductory computational complexity) is required as a prerequisite. The course project involves programming (most likely in Java), thus also programming skills are necessary. Teaching will be given in English. Student projects and term papers are to be written in English.
The students are expected to
finish a
programming project (to be announced later) (50% of the grade)
write a term
paper at the end of the course (50% of the grade)
Project work can be performed individually or in groups. In either case the total amount of work per student is constant. Participation to the classes is not enforced, but strongly encouraged.
During the first two classes of the course anybody can drop out by just sending email to the instructor. At this point the vacant slots will be filled from the pool of students on the waiting list.
 |
|
|
|
|
|
| Graphical Models | |