Abstract | ||
---|---|---|
Computer models are widely used to simulate complex and costly real processes and systems. In the calibration process of the computer model, the calibration parameters are adjusted to fit the model closely to the real observed data. As these calibration parameters are unknown and are estimated based on observed data, it is important to estimate it accurately and account for the estimation uncertainty in the subsequent use of the model. In this paper, we study in detail an empirical Bayes approach for stochastic computer model calibration that accounts for various uncertainties including the calibration parameter uncertainty, and propose an entropy based criterion to improve on the estimation of the calibration parameter. This criterion is also compared with the EIMSPE criterion.
|
Year | DOI | Venue |
---|---|---|
2013 | 10.5555/2675983.2676061 | WSC '13: Winter Simulation Conference
Washington
D.C.
December, 2013 |
Keywords | Field | DocType |
calibration,stochastic processes,parameter estimation | Simulation,Computer science,Algorithm,Stochastic process,Artificial intelligence,Estimation theory,Calibration (statistics),Machine learning,Calibration,Bayes' theorem | Conference |
ISSN | ISBN | Citations |
0891-7736 | 978-1-4799-2077-8 | 4 |
PageRank | References | Authors |
0.49 | 4 | 2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jun Yuan | 1 | 244 | 23.10 |
Szu Hui Ng | 2 | 223 | 21.88 |