Carlos Berenstein, Laveen Kanal, David Lavine
The combination of expert opinions is a problem of broad interest and considerable difficulty. In this paper, we consider a formulation of the problem in which each expert estimates the probability of certain events, and the goal is to produce a single probability distribution which summarizes the various estimates. The study of such combination procedures is called consensus theory. This paper surveys general results on the inherent limitations of the consensus process. In addition, mathematical properties of several linear consensus rules are described.
Keywords: Consensus Theory
PDF Link: /papers/85/p27-berenstein.pdf
AUTHOR = "Carlos Berenstein
and Laveen Kanal and David Lavine",
TITLE = "Consensus Rules",
BOOKTITLE = "Uncertainty in Artificial Intelligence Annual Conference on Uncertainty in Artificial Intelligence (UAI-85)",
PUBLISHER = "Elsevier Science",
ADDRESS = "Amsterdam, NL",
YEAR = "1985",
PAGES = "27--32"