Image 01

3rd International Conference on Belief Functions




A Best Paper award, sponsored by Elsevier, worth 1000 euros will be assigned to the most outstanding technical contribution based on the reviews received by each accepted paper. The final decision will be made by the Belief Functions and Applications Society (BFAS) board.
A Best Student Paper award will be specifically assigned to the best work by a student. This award is sponsored by the International Society of Information Fusion (ISIF). The lead student author will receive a free student registration to attend FUSION 2015.

For the first time we introduced awards at the conference, with the aim of providing recognition to the authors of pieces of work able to generate significant advances in the theory of belief functions, and a sense of identity and common purpose to the community as a whole. A shortlist of 9 candidate papers, two of which student papers, was prepared by the Program Chair on the basis of the reviewers’ scores and assessments. This list was later submitted to the Board of Directors of BFAS for the selection of a Best Paper and a Best Student Paper awards.

The Best Paper Award, sponsored by Elsevier and the International Journal of Approximate Reasoning (IJAR) went to the paper “Evidential Object Recognition based on Information Gain Maximization” by two new members of our community, Thomas Reineking and Kerstin Schill from the University of Bremen, Germany. The paper, which proposes an active object recognition framework based on belief function inference and information gain maximisation, was signalled by the Board as an example of novelty and significant methodological contribution likely to spur further research.

The Best Student Paper Award, sponsored by the International Society for Information Fusion (ISIF), was assigned to the paper “Evidential Logistic Regression for SVM Classifier Calibration” by Ph.D. student Philippe Xu and his advisors Franck Davoine and Thierry Denoeux, from the Université de Technologie de Compiègne, France. The paper proposes an interesting “calibration” method to transform the output of a classifier into a belief function, a significant methodological contribution.