An integrated model of network Data Envelopment Analysis and principal component analysis approach to calculate the efficiency of industrial units (Case study: Stone Industry)

Evaluating the efficiency of industrial units has long been an important issue to find the position of each unit in comparison with others. In this paper, a model for evaluating the efficiency using data envelopment analysis approach is explained in such a way that due to the breadth of input and output criteria, using the principal component analysis approach, data dimensions can also be reduced and the power to distinguish between efficient and inefficient units be increased. Due to lack of attention to the internal structure and also not considering the effective criteria in each department, it is tried to determine the most important criteria involved in each part in the purchasing, production, support and sales sectors. To calculate the efficiency, all the components have been examined as a model of network data envelopment analysis to take into account the effect of all departments and criteria in industrial units' efficiency. In this network, by considering the criteria involved in each of the sub-networks, all effective factors were identified. These criteria are selected based on the SCOR model and the balanced scorecard and also include sustainability criteria. To implement the model, 26 stone factories have been considered. The supply chain network was determined and dimensions of the data were reduced by implementing the principal component analysis approach. Then, by modeling the data envelopment analysis in each of the subnets in GAMS software, the efficiency was calculated. The results show an acceptable difference among industrial units to evaluate those units.