New common set of weights method in black-box and two-stage data envelopment analysis

Data envelopment analysis (DEA) strives to evaluate the production units under their best conditions. DEA flexibility in selecting the appropriate input/output weights always results in unreal and zero weights. Treating decision-making units (DMUs) as black-box regardless of their internal structure...

Full description

Saved in:
Bibliographic Details
Published in:Annals of operations research 2022-02, Vol.309 (1), p.143-162
Main Authors: Kiaei, Hamid, Kazemi Matin, Reza
Format: Article
Language:English
Subjects:
Business and Management
Combinatorics
Data analysis
Data envelopment analysis
Decision making
Industrial equipment
Linear programming
Mathematical programming
Methods
Operations research
Operations Research/Decision Theory
Optimization
Original Research
Performance evaluation
Theory of Computation
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Data envelopment analysis (DEA) strives to evaluate the production units under their best conditions. DEA flexibility in selecting the appropriate input/output weights always results in unreal and zero weights. Treating decision-making units (DMUs) as black-box regardless of their internal structures misleads the DEA performance evaluation. While considering units as a network process, it is more likely to identify more inefficiency sources. This paper suggests using a new common set of weights (CSWs) approach to evaluate the units in both black-box and two-stage structures based on a unified criterion. Indeed, our contribution to this line of research is as follows: Firstly, we improve the model proposed by Kao and Hung (J Oper res Soc 56(10): 1196–1203, 2005) to calculate the CSWs in a linear-based optimization model. Secondly, a new CSWs method is suggested in the two-stage network DEA (NDEA) as multiple objectives fractional programming (MOFP) problem. Thirdly, the MOFP problem is converted into a single objective linear programming problem in the two-stage network case. Finally, an enlightening application is presented.
ISSN:0254-5330
1572-9338
DOI:10.1007/s10479-021-04304-9