Ranking procedures by pairwise comparison using random sets and the imprecise Dirichlet model

Methods for ranking of alternatives or objects by pairwise comparisons using random set theory are proposed in the paper. Efficient algorithms weekly depending on the number of independent sources of data are considered. Methods using the imprecise Dirichlet model are used for obtaining cautious comparison measures when the number of expert judgments is rather small and standard methods of random set theory may give risky results. The methods allow us to overcome some difficulties concerning the conflicting or contradictory sources of data. Various numerical examples illustrate the proposed algorithms and methods.