Multi-criteria inventory classification with reference items

• We evaluate the multi-criteria inventory classification problem as a sorting problem.
• Some reference items, industry or company specific, are assigned in advance to each class by the decision maker.
• We apply two utility function based sorting methods to the problem.
• We conduct several experiments and present our results.

The ABC method is a well-known approach to classify inventory items into ordered categories, such as A, B and C. As emphasized in the literature, it is reasonable to evaluate the inventory classification problem in the multi-criteria context. From this point of view, it corresponds to a sorting problem where categories are ordered. Here, one important issue is that the weights of the criteria and categorization preferences can change from industry to industry. This requires the analysis of the problem in a specific framework where the decision maker (expert)’s preferences are considered. In this study, the preferences of the decision maker are incorporated into the decision making process in terms of reference items into each class. We apply two utility functions based sorting methods to the problem. We perform an experiment and compare results with other algorithms from the literature.