Research Project: Geometry of General Belief Combination
In this project, we start to extend this geometric analysis to several other major combination operators, including Yager's and Dubois' rules, but also the disjunctive operator. The final objective of the research programme is a comparative geometric analysis of combination rules, which would eventually allow us to describe the `cone' of possible future belief states under stronger or weaker assumptions on reliability and independence of sources, associated with conjunctive and disjunctive combination. The bulk of the analysis focusses on standard, normalised belief functions, but we will also consider unnormalised belief functions to provide some preliminary results.
Relevant papers:
  •   Fabio Cuzzolin
    The geometry of uncertainty - The geometry of imprecise probabilities
    Artificial Intelligence: Foundations, Theory, and Algorithms (http://www.springer.com/series/13900), Springer-Verlag, 2018 (in press)
    The geometry of uncertainty
  • Fabio Cuzzolin
    Geometry of General Combination
    Proceedings of the Fifth International Conference on the Theory of Belief Functions (BELIEF 2018)
    Compiegne, France, September 2018