A non-radial directional distance method on classifying inputs and outputs in DEA: Application to banking industry

The original Data Envelopment Analysis (DEA) models have required an assumption that the status of all inputs and outputs be known exactly, whilst we may face a case with some flexible performance measures whose status is unknown. Some classifier approaches have been proposed in order to deal with flexible measures. This contribution develops a new classifier non-radial directional distance method with the aim of taking into account input contraction and output expansion, simultaneously, in the presence of flexible measures. To make the most appropriate decision for flexible measures, we suggest two pessimistic and optimistic approaches from both individual and summative points of view. Finally, a numerical real example in the banking system in the countries of the Visegrad Four (i.e. Czech Republic, Hungary, Poland, and Slovakia) is presented to elaborate applicability of the proposed method. (C) 2017 Elsevier Ltd. All rights reserved.