Dominance-based rough set approach on pairwise comparison tables to decision involving multiple decision makers

Abstract

In this paper, we present a rough set approach to pairwise comparison tables for supporting decisions of multiple decision makers. More precisely, we deal with preference learning from pairwise comparisons, in case of multiple decision makers. Comparing to classical rough set approach, there are three main differences that are the following: we are learning a preference relation, so we have to work with a pairwise comparison table, while the classical rough set approach considers a classification table; we are taking into account a preference order in data, so we have to use the Dominance-based Rough Set Approach (DRSA), while the classical rough set approach based on equivalence relation does not consider such an order; we are taking into account multiple decision makers, while the classical rough set approach considers mostly a single classification decision provided by one decision maker only. © 2011 Springer-Verlag.