Some representative examples of examination questions for URKS:

 

Part 1 (Introduction to Uncertainty – D. De Schreye):

- Explain the notion of Abductive Reasoning. Illustrate with an example
  how this allows you to answer queries in a context with incomplete
  information.

Part 2 (Bayesian - H.Bruyninckx)

Open Book
Make a graphical model of the following knowledge system.  (Max. 3 pages.)
An electric wheelchair is equiped with some sensors (that you can choose)
to observe the environment, and with a sensor on the joystick that the
human driver uses to steer the wheelchair. The goal of the knowledge system
is to estimate the driver intentions of the human, such that it can make
the task of the human lighter.

Make and motivate a graphical model for the following knowledge system.
The system assists the surgeon in the operating room, while he is operating
a patient with a surgical robot consisting of three arms, one of them
equipped with a camera inside the body (to let the surgeon see what he is
doing, but also to track the positions of the other arms inside the body),
and all three arms have markers on their links outside of the body which
are tracked by a set of three camera's. The system is given the surgical
procedure that is executed, and should provide the surgeon with
information about how well he is performing, and give alarms when he is
moving the arms outside of the normal position.

Closed book:
Briefly answer (max 5 lines) the following questions. 
- What is the importance of Bayes' rule for knowledge systems?
- How can one measure information in a probability density function?
- What is the relation between information processing and estimation?
- How can one compare two different models to explain the same data?
- What is the importance of the concept of a Markov Blanket?
- What is the difference between a Markov Random Field and a Bayesian
    Network? 

Part 3 (Possibility theory, Belief functions – M. Nuttin):

- How can you represent "total ignorance" in  the theory of evidence (Theory
  of belief functions). Give a brief explanation. (5 lines).
- Probabilities in probability theory on the one hand, and possibilities in
  possibility theory on the other hand are very different concepts. Explain.
  (max 0.5 page)


Part 4 (Fuzzy systems – M. Nuttin)

- "Generalized modus ponens" plays an important role in fuzzy inference. What
   is it used for? Say you have a fuzzy rule "IF x is A THEN y is B". Explain how to apply
   generalized modus ponens. Max 0.5 page.

- What is "defuzzification" ? Can you think of some defuzzification procedures
   and their advantages and disadvantages? What would you use this procedure
   for ?