Supplier Selection Using Data Envelopment Analysis in the Presence of Imprecise Data

KARSAK E. E., DURSUN M., Ozogul C. O.

15th World Multi-Conference on Systematics, Cybernetics and Informatics (WMSCI 2011), Florida, United States Of America, 19 - 22 July 2011, pp.253-258 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • City: Florida
  • Country: United States Of America
  • Page Numbers: pp.253-258
  • Galatasaray University Affiliated: Yes


This paper aims to present a comprehensive methodology for supplier selection, incorporating both the financial and strategic aspects and the related imprecise as well as exact data into the decision making process. A data envelopment analysis (DEA) model that can take into consideration crisp, ordinal, and fuzzy data is developed for supplier selection. The DEA approach is performed by employing average cost per unit and lead time as the input variables, and number of bills received from the supplier without errors, supplier's experience and supplier reputation as the output variables. The assessment of suppliers versus experience and reputation are represented via ordinal data, while lead time and number of bills received without errors are stated using triangular fuzzy numbers. The proposed framework is illustrated through an example problem for supplier selection.