Kaveh Khalili-Damghani - Department of Industrial Engineering, South-Tehran Branch, Azad University, Tehran, Iran
Hossein Didehkhani - Department of Management & Accounting, Aliabad Katool Branch, Azad University, Golestan, Iran
Mahboobeh Molaiee - Department of Industrial Engineering, South-Tehran Branch, Azad University, Tehran, Iran
Sara J Sharahi - Department of Industrial Engineering, South-Tehran Branch, Azad University, Tehran, Iran

In this paper a hybrid approach based on DEMATEL, ANP, and DEA-based MalmquistProductivity Index is proposed to solve complicated real world Multiple-Criteria Decision Making(MCDM) problems. First, DEMATEL is used to detect the network of interactive criteria through causeand effect relations. Then, the relative importance of the criteria is calculated using ANP method.Finally, a DEA-based Malmquist Productivity Index approach is used to evaluate the productivity ofalternatives with multiple inputs and outputs during several planning periods while the relativeimportance achieved from ANP are also considered as constraints of the system. This approach canhandle complicated real world MCDM problems effectively. In order to provide a clear picture of theperformance of the proposed hybrid approach, it is used at Detection and Governmental PunitiveAgency in all provinces in Iran and the results are analyzed. Important point of this study is theimportance of accuracy in determining the weights of each criterion. The more accurately the weights,closer to the reality are results, and reliability will be higher. Closeness to reality is resulted whenrelations among criteria are determined correctly by experts through DEMATEL and ANP techniques

Data envelopment analysis; DEMATEL; Analytical network process; Malmquist productivity index