An extended of multiple criteria data envelopment analysis models for ratio data
محورهای موضوعی : International Journal of Data Envelopment Analysis
Javad Gerami
1
(
Department of mathematics, Shiraz Branch, Islamic Azad University, Shiraz, Iran
)
کلید واژه: DEA-R, Multi-criteria data envelopment analysis, Discrimination power, Weight dispersion, Pseudo-inefficiency,
چکیده مقاله :
One of the problems of the data envelopment analysis traditional models in the multiple form that is the weights corresponding to certain inputs and outputs are considered zero in the calculation of efficiency and this means that not all input and output components are utilized for the evaluation of efficiency, as some are ignored. The above issue causes the efficiency score of the under evaluation unit not to be calculated correctly. One of the ways to deal with the pseudo-inefficiency is to use data envelopment analysis models with multi-criteria structure. In this regard, we first investigate the models of data envelopment analysis with multi-criteria structure and further, with regard to the ability of the ratio-based data envelopment analysis models, we develop data envelopment analysis models with a multi-criteria structure for ratio data and the feasibility and the bounded condition of the above models and their efficiency intervals are described. By presenting a numerical example, we compare the efficiency scores obtained from the models presented with the previous models and we show that the proposed models can be used to deal with the pseudo-inefficiency and efficiency underestimation. Finally, we present the results.
[1] Andersen, P., & Petersen, N. C. (1993). A procedure for ranking efficient units in data envelopment analysis. Management Science, 39, 1261-1264.
[2] Angiz, M. Z., & Sajedi, M. A. (2010). Improving cross-efficiency evaluation using fuzzy concepts. World Applied Sciences Journal ,16, 1352-1359.
[3] Bal, H., Orkcu, H. H., & Celebioglu, S. (2008). A new method based on the dispersion of weights in data envelopment analysis. Computers & Industrial Engineering, 54, 502-512.
