Research article

A clustering approach of decision making units in data envelopment analysis using axial solutions

Rezamirshafaeia S and keshavarz GF*

Department of Mathematics, Roudsar and Amlash Branch, Islamic Azad University, Roudsar, Iran

*Corresponding Author. Email: atefe.keshavarz@gmail.com

Accepted 28 November 2013

Abstract

The ways of placing decision making units (DMUs) in certain clusters are a subject in statistics, these ways usually are heuristically. The proposed clustering approach in this article considers preferences of DMUs. This study applies data envelopment analysis; DMUs are clustered by solving multi objective production problem (MOLP) and by considering preferences of each DMU at production of each output or consumption of each input. All of DMUs are partitioned into k clusters based on their preference units. The models can be classic models in DEA; in the article clustering DMUs is in the base of CCR envelopment model that reference units play important role. Idea of axial solutions is used for solving MOLPs that considers preferences of DMUs. As result, clustering DMUs is done by optimization solutions that are most preference solution at point of view of decision maker. Also a numerical example is presented and the approach is compared with two other different methods.

Keywords: Clustering approach- DEA- Axial solution-Preferences of functions