Evaluating discriminating power of single-criteria and multi-criteria models towards inventory classification

Abstract

Single-criteria and multi-criteria models both are used with regards to inventory classification. In this paper, we evaluated single-criteria and multi-criteria models in terms of their feasibility in classifying inventory items for a given dataset. We introduced discriminating power test. We used two datasets with lead time as the first criterion. We compared the scores of the models. We also modified ZF model and used descending ranking order criteria constraint to address the infeasibilities. Results show that using criteria in descending order reduces the classification infeasibility. Later, we proposed a probability distribution to find the probability of infeasibility for a given dataset against a number of identical scoring items.