Modelling Agent Judgements for Bayesian Updating

Authors: 
M. West, J. Crosse
Duke University

Jun 30 1989

A robust Bayesian approach to expert, or agent, opinion analysis is developed and applied to various problems of opinion synthesis. A single agent provides a decision maker with probabilistic information partially or completely describing the agent's opinion about a collection of uncertain quantities or events. rules by which the decision maker may update her prior beliefs about related quantities or events of interest are derived. The approach is analogous to that of Genest and Schervish (1985), requiring only a partial specification of the decision maker's prior over the agent's opinion. Results of Genest and Schervish are generalised and applied, and links with alternative modeling approaches are discussed.

Keywords: 

Agent opinion, expert opinion, combining probabilities

Manuscript: 

PDF icon 1989-13.pdf