Aggregation of expert opinions and uncertainty theories

Sébastien Destercke (1, 2) , Didier Dubois (1, 3) , Eric Chojnacki (2)

1 IRIT-ADRIA - Argumentation, Décision, Raisonnement, Incertitude et Apprentissage
2 IRSN - Institut de Radioprotection et de Sûreté Nucléaire
3 CNRS - Centre National de la Recherche Scientifique

The problem of expert opinions representation and aggregation has long been adressed in the only framework of probability theory. Nevertheless, recent years have witnessed many proposals in other uncertainty theories (possibility theory, evidence theory, imprecise probabilities). This paper casts the problem of aggregating expert opinions in a common underlying framework and shows how uncertainty theories fit into this framework. Differences between theories are then emphasized and discussed.

Mots clés: Expert opinions; fusion; possibility theory; belief functions; imprecise probabilities

Domaines: Informatique [cs]; Autre [cs.OH]; Sciences cognitives; Informatique; Sciences de l'ingénieur [physics]; Autre

Sébastien Destercke, Didier Dubois, Eric Chojnacki. Aggregation of expert opinions and uncertainty theories. Rencontres Francophones sur la Logique Floue et ses Applications, 2006, Toulouse, France. pp.295-302. ⟨irsn-00196474⟩