Title: On the Aggregation of Expert Probability Distributions
Abstract: The aggregation of the expert's probability distribution is considered. As opposed to conventional approaches, in which the methods of group decision theory are used, the approach, presented in the paper is based on the probability and information theories. The expert's distributions are considered as conditional distributions. According to this approach the purpose of aggregation is to get the best possible estimation of an unknown probability distribution. In compliance with it, the optimum aggregation procedure is suggested. For the determination of this procedure the informational theoretical criteria and the methods of multiple-criteria decision making are used.