Multi-Item Inventory System Policies Using Statistical Estimates: Negative Binomial Demands (Variance/Mean = 9).
Abstract
In a previous study, MacCormick considered the control of multi-item inventory systems when knowledge of the underlying demand distributions is incomplete. In his simulation study, the distributions actually generating the demands were either Poisson or negative binomial (variance/mean = 3). This report employs negative binomial distributions (variance/mean = 9) to explore the impact of a higher variance of demand on system performance. It also treats the importance of the level of knowledge of the true variance of demand. Whereas much of the theoretical inventory research literature assumes that the probability distributions for demand are exactly known, real life implementation of such systems must inevitably statistically estimate the parameters of the demand distributions. This investigation examines multi-item system behavior of selected operating characteristics as a function of the statistical procedures adopted in implementing stockage rules.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1975
- Accession Number
- ADA015810
Entities
People
- Arthur S. Estey
- Ronald L. Kaufman
Organizations
- Yale University