The Power Approximation: Inventory Policies Based on Limited Demand Information.

Abstract

This investigation examines the problem of managing inventory systems when the probability distributions for demand are incompletely specified. An approximately optimal (s,S) policy rule is derived, requiring knowledge of only the mean and variance of demand. The operating characteristics of the policy are found to be quite close to the characteristics of optimal policies for a wide range of parameter settings. The approximately-optimal policy rule also is examined for the situation in which the decision-maker's knowledge is limited to a sample of previously-realized demands. Policy parameters are revised periodically using a fixed number of past demands to estimate the mean and variance of the demand distribution. The importance of demand information is investigated by also analyzing the process when the mean or variance of demand is known exactly and the other parameter is periodically estimated. In addition, the research examines the accuracy of statistical forecasts that predict the future behavior of the operating characteristics. As a result, an inventory systems designer is apprised of both the costs of imperfect information and the extent of bias in the forecast estimates.

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 1976
Accession Number
ADA026635

Entities

People

  • Richard Ehrhardt

Organizations

  • Yale University

Tags

Communities of Interest

  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Accuracy
  • Inventory
  • Probability
  • Probability Distributions

Readers

  • Logistics and Supply Chain Management.
  • Regression Analysis.
  • Systems Analysis and Design