Evaluation of a 'Large Pop' Filtering Rule for Inventory Management Systems.
Abstract
In this report we evaluate the performance of an inventory filtering rule on computer simulated individual customer orders. The filtering rule identifies a threshold value for which all orders exceeding that value are not filled from existing stock, but rather are specially handled. We present classical statistical outlier rules and other inventory filtering rules and show how they do not perform well in a practical inventory setting. We develop an inventory filtering rule, and test its performance on seven different customer order distributions that resemble distributions we have seen in practice. We shown that for practical inventory applications, our filtering rule statistically outperforms other models currently in the literature. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1983
- Accession Number
- ADA124686
Entities
People
- Douglas Blazer
Organizations
- University of North Carolina at Chapel Hill