Some guidelines and guarantees for common random numbers
Article Abstract:
Common random numbers (CRN) is a popular, easily applied variance reduction method that has been widely adopted as part of the simulation techniques used to analyze different stochastic systems. While CRN is known to be effective for many kinds of models, its use is considered optimal for only a limited number of model classes. The application of CRN in three situations is discussed in relation to how the effectiveness and optimality of CRN is influenced by continuity factors and monotonicity. These three situations were designed to correspond to three common CRN applications, such as sensitivity analysis, structural comparisons and distributional analysis.
Publication Name: Management Science
Subject: Business, general
ISSN: 0025-1909
Year: 1992
User Contributions:
Comment about this article or add new information about this topic:
Reducing the congestion in a class of job shops
Article Abstract:
Production planning is analyzed for job shops where all work stations have the same number of workers. The objective is to optimize labor and storage capacity for the whole shop as well as for individual stations. Two problems were addressed: station work load and workers-per-station, using loading and assignment vectors. Methodology involved majorization, open queuing networks, likelihood of ratio ordering, and stochastic ordering. It is demonstrated that it is better to load stations to their equilibrium point, or to assign faster workers more work. It is assumed that all stations have an equal number of workers.
Publication Name: Management Science
Subject: Business, general
ISSN: 0025-1909
Year: 1987
User Contributions:
Comment about this article or add new information about this topic:
Optimal server allocation in a system of multi-server stations
Article Abstract:
The optimal configuration of workers in a system of multiple stations is analyzed. The system is of the closed queuing network type, with the productivity rate of each station being an increasing concave function of the number of tasks performed there. This is demonstrated through sample path comparisons of certain constructed stochastic processes. Using the results, it is possible to determine the best allocation of workers in multi-worker stations with limited buffers, as well as the best arrangement of workers and buffers in single-station networks.
Publication Name: Management Science
Subject: Business, general
ISSN: 0025-1909
Year: 1987
User Contributions:
Comment about this article or add new information about this topic:
- Abstracts: Optimal and heuristic procedures for component lot-splitting in multi-stage manufacturing systems. Piecewise loglinear estimation of efficient production surfaces
- Abstracts: JIT: cost accounting and cost management issues. How cost management systems can support the JIT philosophy. Eleven key issues in manufacturing accounting
- Abstracts: Share influence and organizational behavior: a meta-analysis of situational variables expected to moderate participation-outcome relationships
- Abstracts: Accountants as business partners. How challenging should profit budget targets be? Early warning systems
- Abstracts: Delegated information gathering decisions. Information asymmetry, diversification, and cost of capital. Commitments and disclosure in oligopolies