Multi-Criteria Inventory Classification (MCIC) groups inventory items with respect to several criteria, in order to facilitate their management. This paper introduces a new hybrid method based on AHP and the K-means algorithm. On benchmarking data, it provides a clearly higher clustering validity index than previous sorting methods. However, as with previous methods, it is a full compensatory method. This means that an item scoring badly on one or more key criteria may still be placed in the best class because these bad scores are compensated. In order to prevent these hidden bad scores, a new variant method is introduced: AHP-K-Veto. The sorting is performed on each single criterion, where a veto system prevents an item evaluated as high/bad on at least one criterion to be top/bottom ranked in the global aggregation. This veto system is an assurance against hidden problems but slightly worsens the clustering validity index.
- Multiple criteria analysis
- Assignment, Inventory