Adaptive Analysis of Experts’ Statements in Pattern Recognition

Abstract

In this paper, we consider some problems related to forming a consensus of experts' statements for the case of probabilistic forecasting of qualitative feature. We assume that decision rule is constructed on the base of analysis of empirical information represented in the form of probabilistic statements from several experts. The criterion of a quality of experts’ statements is suggested. The method of forming of united expert decision rule using distances / similarities between multidimensional sets in heterogeneous feature space is proposed.