Title: Multi-criteria ABC inventory classification using DEA-discriminant analysis to predict group membership of new items

Authors: Mohammad Tavassoli; Gholam Reza Faramarzi; Reza Farzipoor Saen

Addresses: Department of Industrial Management, Faculty of Management and Accounting, Karaj Branch, Islamic Azad University, Karaj, P.O. Box 31485-313, Iran ' Department of Industrial Management, Faculty of Management and Accounting, Karaj Branch, Islamic Azad University, Karaj, P.O. Box 31485-313, Iran ' Department of Industrial Management, Faculty of Management and Accounting, Karaj Branch, Islamic Azad University, Karaj, P.O. Box 31485-313, Iran

Abstract: Inventory management plays a significant role in organisations' success or failure. ABC inventory classification is one of the most popular methods which are regularly applied in inventory management. Correct clustering of inventory items is an important issue of inventory management. The 'annual cost' is an important factor in most of previous studies which applied ABC inventory classification. Each item which has higher annual cost is placed in class A. This paper shows that other factors have significant role for classifying inventory items. We use data envelopment analysis (DEA) to classify inventory items into three groups as A, B, or C in the presence of weight restrictions. Weight restrictions allow for the integration of managerial preferences in terms of relative importance of various factors. Then, to predict group membership of new items, the DEA is incorporated with discriminant analysis (DA). To demonstrate applicability of proposed approach a case study is presented.

Keywords: ABC classification; data envelopment analysis; DEA; discriminant analysis; multicriteria inventory classification; group membership; new items; inventory management; managerial preferences; case study.

DOI: 10.1504/IJAMS.2014.060904

International Journal of Applied Management Science, 2014 Vol.6 No.2, pp.171 - 189

Published online: 02 Jul 2014 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article