Use of multi-criteria decision analysis in fuzzy network DEA models

Data Envelopment Analysis (DEA) is a popular tool for examining the efficiency of production units in transforming a set of inputs into outputs; however, traditional models consider them to be closed, black-box systems. Network DEA (NDEA) allows one to capture the inner structure as a series of interacting processes, making for a more precise model of complex systems. In an effort to further increase the accuracy of real world representation in models, several directions of research into NDEA have been developed. One of them replaces crisp data with fuzzy values to better capture the imprecise nature of most complex systems. Elsewhere, models that use a number of Multi-criteria Decision Analysis (MCDA) methods allow managers to put restrictions on allowable input/output weights, to rank efficient units, or to solve other scenarios for which traditional models are ill-equipped. This paper aims to first provide an overview of existing models utilizing both aforementioned approaches. Subsequently, further ways in which both approaches could be combined to yield brand new methods are examined. Advantages and limitations of such endeavor, as well as proposed practical applications of this combined technique, follow.