Take a look at the new version of this page: [ beta version ]. Tell us what you think.

ABSTRACT

One goal in simulation experimentation is to identify which input parameters most significantly influence the mean of simulation output. Another goal is to obtain good parameter estimates for a response model that quantifies how the mean output depends on influential input parameters. The majority of experimental design techniques focus on either one goal or the other. This paper uses a design criterion for follow-up experiments that jointly identifies the important parameters and reduces the variance of parameter estimates. The criterion is entropy-based, and is applied to a critical care facility simulation.

REFERENCES

• 1: Emile Aarts , Jan Korst, Simulated annealing and Boltzmann machines: a stochastic approach to combinatorial optimization and neural computing, John Wiley & Sons, Inc., New York, NY, 1989
• 2: Russell R. Barton, Simulation metamodels, Proceedings of the 30th conference on Winter simulation, p.167-176, December 13-16, 1998, Washington, D.C., United States
• 3: Bernardo, J. 1979. Expected information as expected utility. Annals of Statistics 7:686--690.
• 4: Bernardo, J. M., and A. F. M. Smith. 1994. Bayesian theory. Chichester, UK: Wiley.
• 5: Borth, D. 1975. A total entropy criterion for the dual problem of model discrimination and parameter estimation. Journal of the Royal Statistical Society, Series B 37 (1): 77--87.
• 6: Box, G., and W. Hill. 1967. Discrimination among mechanistic models. Technometrics 9 (1): 57--71.
• 7: Cheng, R. C. H., and W. Holland. 1997. Sensitivity of computer simulation experiments to errors in input data. Journal on Statistical Compututing and Simulation 57:219--241.
• 8: DeGroot, M. 1962. Uncertainty, information and sequential experiments. Annals of Statistics 33:404--419.
• 9: Hill, P. 1978. A review of experimental design procedures for regression model discrimination. Technometrics 20 (1): 15--21.
• 10: Hill, W., and W. Hunter. 1969. A note on designs for model discrimination: Variance unknown case. Technometrics 11 (1): 396--400.
• 11: Hill, W., W. Hunter, and D. Wichern. 1968. A joint design criterion for the dual problem of model discrimination and parameter estimation. Technometrics 10 (1): 145--160.
• 12: Sheldon H. Jacobson , Arnold H. Buss , Lee W. Schruben, Driving frequency selection for frequency domain simulation experiments, Operations Research, v.39 n.6, p.917-924, Nov.-Dec. 1991 [doi>10.1287/opre.39.6.917]
• 13: Johnson, M., and C. Nachtsheim. 1983. Some guidelines for constructing exact d-optimal designs on convex design spaces. Technometrics 25 (3): 271--277.
• 14: Kleijnen, J. 1996. Experimental design for sensitivity analysis, optimization, and validation of simulation models. In Handbook of Simulation, ed. J. Banks. New York: John Wiley & Sons, Inc.
• 15: Averill M. Law , David M. Kelton, Simulation Modeling and Analysis, McGraw-Hill Higher Education, 1999
• 16: Lindley, D. V. 1956. On a measure of the information provided by an experiment. Annals of Mathematical Statistics 27:986--1005.
• 17: Madigan, D., and J. York. 1995. Bayesian graphical models for discrete data. International Statistical Review 63 (2): 215--232.
• 18: R. Daniel Meyer , David M. Steinberg , George Box, Follow-up designs to resolve confounding in multifactor experiments, Technometrics, v.38 n.4, p.303-313, Nov. 1996 [doi>10.2307/1271297]
• 19: Douglas J. Morrice , Lee W. Schruben, Simulation factor screening using harmonic analysis, Management Science, v.39 n.12, p.1459-1476, Dec. 1993 [doi>10.1287/mnsc.39.12.1459]
• 20: Ng, S. H. 2001. Sensitivity and uncertainty analysis of complex simulation models. Ph. D. thesis, The University of Michigan, Ann Arbor, MI. Dept. of Industrial and Operations Engineering.
• 21: Szu Hui Ng , Stephen E. Chick, Reducing input parameter uncertainty for simulations, Proceedings of the 33nd conference on Winter simulation, December 09-12, 2001, Arlington, Virginia
• 22: Ng, S.-H., and S. E. Chick. 2002. A Bayesian follow up design criterion for the dual objective of model discrimination and parameter estimation. Technical report, INSEAD.
• 23: Raftery, A. E., D. Madigan, and J. A. Hoeting. 1997. Bayesian model averaging for linear regression models. Journal of the American Statistical Association 92 (437): 179--191.
• 24: Susan M. Sanchez, A robust design tutorial, Proceedings of the 26th conference on Winter simulation, p.106-113, December 11-14, 1994, Orlando, Florida, United States
• 25: Schruben, L. W., and B. H. Margolin. 1978. Pseudorandom number assignment in statistically designed simulation and distribution sampling experiments. Journal of the American Statistical Association 73 (363): 504--525.
• 26: Leyuan Shi , Sigurdur Ólafsson, Nested Partitions Method for Global Optimization, Operations Research, v.48 n.3, p.390-407, May 2000 [doi>10.1287/opre.48.3.390.12436]
• 27: Smith, A. F. M., and I. Verdinelli. 1980. A note on Bayesian design for inference using hierarchical linear model. Biometrika 67 (3): 613--619.