Examinations for Probabilistic Reasoning in AI (COMP-526)
Winter 2004
The second exam will cover selected topics from lectures 17-30
(including the lecture we have next Wednesday, March 24). List of
topics:
-
Expectation maximization (EM)
-
Hidden Markov models (HMM)
-
Computing expected utilities for actions. Simple decision graphs (1-2
actions).
-
Markov Decision processes: definition, formulating problems as MDPs.
-
Value functions, dynamic programming
-
Learning in MDPs
-
Defining an appropriate probabilistic representation for a problem
Topics that are not on the exam:
-
Kalman filters
-
General inference algorithm for decision graphs
-
Lectures from March 26 onwards.
Possible dates for the exam are:
-
Monday, March 29, in the afternoon (note that Doina will be away that day!)
-
Tuesday April 8, in the afternoon
-
Tuesday, April 13, in the afternoon
Note that the list of topics does not depend on the exam date
Prof. Doina PRECUP
Last modified: Wed Mar 18 21:45:59 EST 2004