Instructor: Thierry Denoeux

Description: This course presents the main formalisms for uncertainty representation with emphasis on the theory of belief functions, which generalizes both probability theory and the set-membership approach (inclusing interval analysis). Numerous examples taken from different domains (Artificial Intelligence, multi-sensor fusion, numerical modeling) will illustrate the theoretical notions.

Programme (Fall 2013)

1. 11 September: Introduction to uncertainty
2. 18 September: Belief functions: representation of evidence
3. 25 September: Belief functions: combination of evidence
4. 2 October: Refinement and coarsening of a frame of discernment
5. 5 October:
6. 9 October: Applications to pattern recognition and information fusion
7. 12 October: Presentation of research papers by students