Why Dempster's fusion rule is not a generalization of Bayes fusion rule

Dezert, J. ; Tchamova, A. ; Han, D. ; Tacnet, J.M.

Type de document
Communication scientifique avec actes
Langue
Anglais
Affiliation de l'auteur
ONERA PALAISEAU FRA ; BULGARIAN ACADEMY OF SCIENCES SOFIA BGR ; XI'AN JIAOTONG UNIVERSITY CHN ; IRSTEA GRENOBLE UR ETGR FRA
Année
2013
Résumé / Abstract
In this paper, we analyze Bayes fusion rule in details from a fusion standpoint, as well as the emblematic Dempster's rule of combination introduced by Shafer in his Mathematical Theory of evidence based on belief functions. We propose a new interesting formulation of Bayes rule and point out some of its properties. A deep analysis of the compatibility of Dempster's fusion rule with Bayes fusion rule is done. We show that Dempster's rule is compatible with Bayes fusion rule only in the very particular case where the basic belief assignments (bba's) to combine are Bayesian, and when the prior information is modeled either by a uniform probability measure, or by a vacuous bba. We show clearly that Dempster's rule becomes incompatible with Bayes rule in the more general case where the prior is truly informative (not uniform, nor vacuous). Consequently, this paper proves that Dempster's rule is not a generalization of Bayes fusion rule.
Congrès
16th International Conference on Information Fusion, 09/07/2013 - 12/07/2013, Istanbul, TUR

puce  Accés à la notice sur le site Irstea Publications / Display bibliographic record on Irstea Publications website

  Liste complète des notices de CemOA