Hybrid cluster and data envelopment analysis with interval data

Data Envelope Analysis (DEA) is an approach to estimating the relative efficiency of Decision-Making Units (DMUs). Several studies have been conducted in order to prioritize efficient units, and some useful models, such as Cross-Efficiency Matrix (CEM), have been presented. Besides, a number of DEA models with interval data have been developed, and ranking DMUs with such data has been carried out. However, presenting an obtained crisp data derived from interval data is a critical problem; hence, many researches have been conducted so as to compute the weights and average of the interval data. This paper proposes a new algorithm to find highly suitable weights by applying a data mining approach of DMU's data. For this purpose, clustering and a pair-wise comparison matrix were employed to estimate the given relative efficiency of CEM. Results indicate that there is a meaningful difference between efficiency of DMUs with the lower bound and that of DMUs with the upper bound. (C) 2018 Sharif University of Technology. All rights reserved.