Forecasting efficiency of green suppliers by dynamic data envelopment analysis and artificial neural networks

Abstract

Traditional models of data envelopment analysis (DEA) and dynamic DEA cannot forecast future efficiency of decision making units (DMUs). In other words, all models of DEA and dynamic DEA evaluate and rank DMUs based on past performance. This paper opens a new perspective to realm of DEA as it proposes a transition from previous supervising models to a future planning approach which contains novel contributions. For the first time, artificial neural networks (ANN) are combined with dynamic DEA to forecast future efficiency of DMUs (green suppliers). To this end, firstly, we forecast inputs, outputs, and links of the green suppliers using ANN. Then, the forecasted data derived from ANN are used in dynamic DEA. Dynamic DEA evaluates green suppliers in past, present, and future periods, simultaneously. Our proposed approach has helpful outcomes for decision makers. A case study demonstrates applicability of the proposed approach.