Jointly Robust Prior for Gaussian Stochastic Process in Emulation, Calibration and Variable Selection

被引:30
作者
Gu, Mengyang [1 ]
机构
[1] Johns Hopkins Univ, Dept Appl Math & Stat, Baltimore, MD 21218 USA
来源
BAYESIAN ANALYSIS | 2019年 / 14卷 / 03期
关键词
computer model; posterior propriety; reference prior; tail rate; OBJECTIVE BAYESIAN-ANALYSIS; COMPUTER EXPERIMENTS; SENSITIVITY-ANALYSIS; MODELS;
D O I
10.1214/18-BA1133
中图分类号
O1 [数学];
学科分类号
0701 ; 070101 ;
摘要
Gaussian stochastic process (GaSP) has been widely used in two fundamental problems in uncertainty quantification, namely the emulation and calibration of mathematical models. Some objective priors, such as the reference prior, are studied in the context of emulating (approximating) computationally expensive mathematical models. In this work, we introduce a new class of priors, called the jointly robust prior, for both the emulation and calibration. This prior is designed to maintain various advantages from the reference prior. In emulation, the jointly robust prior has an appropriate tail decay rate as the reference prior, and is computationally simpler than the reference prior in parameter estimation. Moreover, the marginal posterior mode estimation with the jointly robust prior can separate the influential and inert inputs in mathematical models, while the reference prior does not have this property. We establish the posterior propriety for a large class of priors in calibration, including the reference prior and jointly robust prior in general scenarios, but the jointly robust prior is preferred because the calibrated mathematical model typically predicts the reality well. The jointly robust prior is used as the default prior in two new R packages, called "RobustGaSP" and "RobustCalibration", available on CRAN for emulation and calibration, respectively.
引用
收藏
页码:857 / 885
页数:29
相关论文
共 40 条
  • [1] [Anonymous], THESIS
  • [2] [Anonymous], THESIS
  • [3] [Anonymous], ARXIV180101874
  • [4] [Anonymous], THESIS
  • [5] [Anonymous], SENSITIVITY PACKAGE
  • [6] Quantification of Model Uncertainty: Calibration, Model Discrepancy, and Identifiability
    Arendt, Paul D.
    Apley, Daniel W.
    Chen, Wei
    [J]. JOURNAL OF MECHANICAL DESIGN, 2012, 134 (10)
  • [7] Diagnostics for Gaussian Process Emulators
    Bastos, Leonardo S.
    O'Hagan, Anthony
    [J]. TECHNOMETRICS, 2009, 51 (04) : 425 - 438
  • [8] Using Statistical and Computer Models to Quantify Volcanic Hazards
    Bayarri, M. J.
    Berger, James O.
    Calder, Eliza S.
    Dalbey, Keith
    Lunagomez, Simon
    Patra, Abani K.
    Pitman, E. Bruce
    Spiller, Elaine T.
    Wolpert, Robert L.
    [J]. TECHNOMETRICS, 2009, 51 (04) : 402 - 413
  • [9] A framework for validation of computer models
    Bayarri, Maria J.
    Berger, James O.
    Paulo, Rui
    Sacks, Jerry
    Cafeo, John A.
    Cavendish, James
    Lin, Chin-Hsu
    Tu, Jian
    [J]. TECHNOMETRICS, 2007, 49 (02) : 138 - 154
  • [10] Berger James O, 1998, Sankhya: The Indian Journal of Statistics, Series A, P307