Clustering Decision Making Units in Data Envelopment Analysis by the Common Set of Weights

3 Abstract: Clustering is the process of grouping data objects into sets of disjoint classes called clusters so that objects within a class are highly similar to one another and dissimilar to the objects in other classes. In this paper, we propose a new clustering method involving data envelopment analysis (DEA) that clusters decision making units (DMUs) to Kpartitions. Therefore, we introduce two virtual decision making units called the positive ideal DMU and the negative ideal DMU , then we define K-2 other virtual decision making units by these two DMUs; after that we use a process of K stages, in each stage of which we define the benchmark line and calculate the distance from each real DMU to this benchmark line; then we will introduce the criterion for clustering DMUs. We apply this clustering method to a numerical example.