Home > Research > Publications & Outputs > Reducing Simulation Input-Model Risk via Input ...

Electronic data

  • NelsonWanEtAlIJoCAAM

    Accepted author manuscript, 565 KB, PDF document

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

Links

Text available via DOI:

View graph of relations

Reducing Simulation Input-Model Risk via Input Model Averaging

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Reducing Simulation Input-Model Risk via Input Model Averaging. / Nelson, Barry; Wan, Alan T. K.; Zou, Guohua et al.
In: INFORMS Journal on Computing, Vol. 33, No. 2, 30.04.2021, p. 672-684.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Nelson, B, Wan, ATK, Zou, G, Zhang, X & Jiang, X 2021, 'Reducing Simulation Input-Model Risk via Input Model Averaging', INFORMS Journal on Computing, vol. 33, no. 2, pp. 672-684. https://doi.org/10.1287/ijoc.2020.0994

APA

Nelson, B., Wan, A. T. K., Zou, G., Zhang, X., & Jiang, X. (2021). Reducing Simulation Input-Model Risk via Input Model Averaging. INFORMS Journal on Computing, 33(2), 672-684. https://doi.org/10.1287/ijoc.2020.0994

Vancouver

Nelson B, Wan ATK, Zou G, Zhang X, Jiang X. Reducing Simulation Input-Model Risk via Input Model Averaging. INFORMS Journal on Computing. 2021 Apr 30;33(2):672-684. Epub 2020 Oct 6. doi: 10.1287/ijoc.2020.0994

Author

Nelson, Barry ; Wan, Alan T. K. ; Zou, Guohua et al. / Reducing Simulation Input-Model Risk via Input Model Averaging. In: INFORMS Journal on Computing. 2021 ; Vol. 33, No. 2. pp. 672-684.

Bibtex

@article{e18444a214a84976b10033c2aae61b25,
title = "Reducing Simulation Input-Model Risk via Input Model Averaging",
abstract = "Abstract. Input uncertainty is an aspect of simulation model risk that arises when the driving input distributions are derived or “fit” to real-world, historical data. Although there has been significant progress on quantifying and hedging against input uncertainty, there has been no direct attempt to reduce it via better input modeling. The meaning of “better” depends on the context and the objective: Our context is when (a) there are one or more families of parametric distributions that are plausible choices; (b) the real-world historical data are not expected to perfectly conform to any of them; and (c) our primary goal is to obtain higher-fidelity simulation output rather than to discover the “true” distribution. In this paper, we show that frequentist model averaging can be an effective way to create input models that better represent the true, unknown input distribution, thereby reducing model risk. Input model averaging builds from standard input modeling practice, is not computationally burdensome, requires no change in how the simulation is executed nor any follow-up experiments, and is available on the Comprehensive R Archive Network CRAN). We provide theoretical and empirical support for our approach.",
keywords = "input modeling, stochastic simulation, input uncertainty",
author = "Barry Nelson and Wan, {Alan T. K.} and Guohua Zou and Xinyu Zhang and Xi Jiang",
note = "{\textcopyright} 2020 INFORMS",
year = "2021",
month = apr,
day = "30",
doi = "10.1287/ijoc.2020.0994",
language = "English",
volume = "33",
pages = "672--684",
journal = "INFORMS Journal on Computing",
issn = "1091-9856",
publisher = "INFORMS Inst.for Operations Res.and the Management Sciences",
number = "2",

}

RIS

TY - JOUR

T1 - Reducing Simulation Input-Model Risk via Input Model Averaging

AU - Nelson, Barry

AU - Wan, Alan T. K.

AU - Zou, Guohua

AU - Zhang, Xinyu

AU - Jiang, Xi

N1 - © 2020 INFORMS

PY - 2021/4/30

Y1 - 2021/4/30

N2 - Abstract. Input uncertainty is an aspect of simulation model risk that arises when the driving input distributions are derived or “fit” to real-world, historical data. Although there has been significant progress on quantifying and hedging against input uncertainty, there has been no direct attempt to reduce it via better input modeling. The meaning of “better” depends on the context and the objective: Our context is when (a) there are one or more families of parametric distributions that are plausible choices; (b) the real-world historical data are not expected to perfectly conform to any of them; and (c) our primary goal is to obtain higher-fidelity simulation output rather than to discover the “true” distribution. In this paper, we show that frequentist model averaging can be an effective way to create input models that better represent the true, unknown input distribution, thereby reducing model risk. Input model averaging builds from standard input modeling practice, is not computationally burdensome, requires no change in how the simulation is executed nor any follow-up experiments, and is available on the Comprehensive R Archive Network CRAN). We provide theoretical and empirical support for our approach.

AB - Abstract. Input uncertainty is an aspect of simulation model risk that arises when the driving input distributions are derived or “fit” to real-world, historical data. Although there has been significant progress on quantifying and hedging against input uncertainty, there has been no direct attempt to reduce it via better input modeling. The meaning of “better” depends on the context and the objective: Our context is when (a) there are one or more families of parametric distributions that are plausible choices; (b) the real-world historical data are not expected to perfectly conform to any of them; and (c) our primary goal is to obtain higher-fidelity simulation output rather than to discover the “true” distribution. In this paper, we show that frequentist model averaging can be an effective way to create input models that better represent the true, unknown input distribution, thereby reducing model risk. Input model averaging builds from standard input modeling practice, is not computationally burdensome, requires no change in how the simulation is executed nor any follow-up experiments, and is available on the Comprehensive R Archive Network CRAN). We provide theoretical and empirical support for our approach.

KW - input modeling

KW - stochastic simulation

KW - input uncertainty

U2 - 10.1287/ijoc.2020.0994

DO - 10.1287/ijoc.2020.0994

M3 - Journal article

VL - 33

SP - 672

EP - 684

JO - INFORMS Journal on Computing

JF - INFORMS Journal on Computing

SN - 1091-9856

IS - 2

ER -