An integrated weighting and ranking model based on entropy, DEA and PCA considering two aggregation approaches for resilient supplier selection problem

R. Davoudabadi | Seyed Meysam Mousavi | E. Sharifi

---

URL :   http://research.shahed.ac.ir/WSR/WebPages/Report/PaperView.aspx?PaperID=137376
Date :  2020/01/12
Publish in :    Journal of Computational Science

Link :  https://www.sciencedirect.com/science/article/abs/pii/S1877750319305113
Keywords :Efficiency measurement model, Data envelopment analysis Principal components analysis (PCA), Entropy, First and last aggregation approaches, Resilient supplier selection problem

Abstract :

---

Choosing a suitable supplier is one of the crucial issues in supply chain management, gaining much attention in the light of disrupted balances between cost-time-quality interactions in developed/developing countries in these past years. In the meantime, the supplier selection with considering resilient criteria is the new topic in this area. Resilient may be defined as sufficient flexibility in counteracting fluctuations. This study presents a new integrated efficiency measurement model combining statistical techniques, decision making, and mathematical programming for resilient supplier analysis. Also, a new combination of two methods of first aggregation and last aggregation is developed to take advantage of both. Methodologically, this paper applies principal components analysis (PCA) to reduce the dimensions and the correlation between the criteria. The PCA method is utilized to decrease the dimensions and the correlation between the criteria. Besides, data envelopment analysis (DEA) is employed to determine the weights of the criteria and ranking the suppliers. Weights of the criteria are established using the DEA method, the entropy, and judgments of decision-makers (DMs) simultaneously. A case study of resilient supplier selection problems is resolved and compared with existing methods to prove the performance of the proposed model. As a way of calculating the weight of the criteria, this study analyzes the performance of the proposed method and compare it with the first and last aggregation approaches. Then, the results of the study are reported.