Applying fuzzy preference relations to ranking decision making units in Data Envelopment Analysis

Abstract

In Data Envelopment Analysis (DEA) models, often more than one efficient unit is available. To rank these units, many methods have been proposed. In recent years, using DEA and preference relation, many methods for ranking full the efficient units is presented. The fundamental problem, exist alternative Optimal solutions in DEA model have been suggested. Exist alternative Optimal solutions may assign more than one rank to a DMU. In this article, a new model of joint weights in DEA to obtain priority matrix (paired comparison matrix) is suggested. Then the fuzzy preference matrix can be built. More to using Eigen Values method in the AHP, the priority vector is obtained for the ranking DMUs. Also a way to produce unique optimal solutions for joint weight model proposed is provided. Advantage of unique optimal weights is that only a single ranking obtained for each DMU with a solution by each decision maker and any software.