Improving Global Efficiency of Organizations by Merging Inefficient DMUs: an ANN-DEA Integrated Approach

We suppose an organization where n homogeneous Decision Making Units (DMUs) areoperating to accomplish the objective of the organization. It is desired to improve the globalefficiency of the organization by evaluating the efficiency of individual DMUs. For thispurpose, sometimes it is recommended to merge the least efficient DMUs. In this paper, weprovide a hybrid algorithm using Data Envelopment Analysis (DEA) and Artificial NeuralNetworks (ANNs) to prescribe the merger for inefficient DMUs, also introduce differentcriteria to show the global efficiency of the organization. An application of the proposedmethodology has been studied in an Iranian commercial bank.

Keywords:

Data Envelopment Analysis (DEA) , Artificial Neural Networks (ANNs) , Global Efficiency , Return to Scale (RTS)