The Annals of Applied Statistics

Composite Gaussian process models for emulating expensive functions

Shan Ba and V. Roshan Joseph

Full-text: Open access

Enhanced PDF (388 KB)

Abstract

A new type of nonstationary Gaussian process model is developed for approximating computationally expensive functions. The new model is a composite of two Gaussian processes, where the first one captures the smooth global trend and the second one models local details. The new predictor also incorporates a flexible variance model, which makes it more capable of approximating surfaces with varying volatility. Compared to the commonly used stationary Gaussian process model, the new predictor is numerically more stable and can more accurately approximate complex surfaces when the experimental design is sparse. In addition, the new model can also improve the prediction intervals by quantifying the change of local variability associated with the response. Advantages of the new predictor are demonstrated using several examples.

Article information

Source
Ann. Appl. Stat., Volume 6, Number 4 (2012), 1838-1860.

Dates
First available in Project Euclid: 27 December 2012

Permanent link to this document
https://projecteuclid.org/euclid.aoas/1356629062

Digital Object Identifier
doi:10.1214/12-AOAS570

Mathematical Reviews number (MathSciNet)
MR3058685

Zentralblatt MATH identifier
1257.62089

Keywords
Computer experiments functional approximation kriging nugget nonstationary Gaussian process

Citation

Ba, Shan; Joseph, V. Roshan. Composite Gaussian process models for emulating expensive functions. Ann. Appl. Stat. 6 (2012), no. 4, 1838--1860. doi:10.1214/12-AOAS570. https://projecteuclid.org/euclid.aoas/1356629062


Export citation

References

  • Ababou, R., Bagtzoglou, A. C. and Wood, E. F. (1994). On the condition number of covariance matrices in kriging, estimation, and simulation of random fields. Math. Geol. 26 99–133.
    Mathematical Reviews (MathSciNet): MR1256401
    Zentralblatt MATH: 0970.86543
    Digital Object Identifier: doi:10.1007/BF02065878
  • Anderes, E. B. and Stein, M. L. (2008). Estimating deformations of isotropic Gaussian random fields on the plane. Ann. Statist. 36 719–741.
    Mathematical Reviews (MathSciNet): MR2396813
    Zentralblatt MATH: 1133.62077
    Digital Object Identifier: doi:10.1214/009053607000000893
    Project Euclid: euclid.aos/1205420517
  • Ankenman, B., Nelson, B. L. and Staum, J. (2010). Stochastic kriging for simulation metamodeling. Oper. Res. 58 371–382.
    Mathematical Reviews (MathSciNet): MR2674803
    Zentralblatt MATH: 05978643
    Digital Object Identifier: doi:10.1287/opre.1090.0754
  • Ba, S. and Joseph, V. R. (2011). Multi-layer designs for computer experiments. J. Amer. Statist. Assoc. 106 1139–1149.
    Mathematical Reviews (MathSciNet): MR2894770
    Zentralblatt MATH: 1229.62102
    Digital Object Identifier: doi:10.1198/jasa.2011.tm10229
  • Banerjee, S., Charlin, B. P. and Gelfand, A. E. (2003). Hierarchical Modeling and Analysis for Spatial Data. Chapman & Hall/CRC, Boca Raton, FL.
    Zentralblatt MATH: 1053.62105
  • Cressie, N. A. C. (1991). Statistics for Spatial Data. Wiley, New York.
    Mathematical Reviews (MathSciNet): MR1127423
    Zentralblatt MATH: 0799.62002
  • Currin, C., Mitchell, T., Morris, M. and Ylvisaker, D. (1991). Bayesian prediction of deterministic functions, with applications to the design and analysis of computer experiments. J. Amer. Statist. Assoc. 86 953–963.
    Mathematical Reviews (MathSciNet): MR1146343
    Digital Object Identifier: doi:10.1080/01621459.1991.10475138
  • Fang, K.-T., Li, R. and Sudjianto, A. (2006). Design and Modeling for Computer Experiments. Chapman & Hall/CRC, Boca Raton, FL.
    Mathematical Reviews (MathSciNet): MR2223960
    Zentralblatt MATH: 1093.62117
  • Gneiting, T. and Raftery, A. E. (2007). Strictly proper scoring rules, prediction, and estimation. J. Amer. Statist. Assoc. 102 359–378.
    Mathematical Reviews (MathSciNet): MR2345548
    Zentralblatt MATH: 05191574
    Digital Object Identifier: doi:10.1198/016214506000001437
  • Gramacy, R. B. and Lee, H. K. H. (2008). Bayesian treed Gaussian process models with an application to computer modeling. J. Amer. Statist. Assoc. 103 1119–1130.
    Mathematical Reviews (MathSciNet): MR2528830
    Zentralblatt MATH: 1205.62218
    Digital Object Identifier: doi:10.1198/016214508000000689
  • Gramacy, R. B. and Lee, H. K. H. (2012). Cases for the nugget in modeling computer experiments. Statist. Comput. 22 713–722.
    Mathematical Reviews (MathSciNet): MR2909617
    Zentralblatt MATH: 1252.62098
    Digital Object Identifier: doi:10.1007/s11222-010-9224-x
  • Haaland, B. and Qian, P. Z. G. (2011). Accurate emulators for large-scale computer experiments. Ann. Statist. 39 2974–3002.
  • Higdon, D. M., Swall, J. and Kern, J. (1999). Non-stationary spatial modeling. In Bayesian Statistics 6, Proceedings of the Sixth Valencia International Meeting 761–768. Oxford Univ. Press, London.
    Zentralblatt MATH: 1005.00041
  • Huang, W., Wang, K., Breidt, F. J. and Davis, R. A. (2011). A class of stochastic volatility models for environmental applications. J. Time Series Anal. 32 364–377.
    Mathematical Reviews (MathSciNet): MR2841790
    Digital Object Identifier: doi:10.1111/j.1467-9892.2011.00735.x
  • Joseph, V. R. (2006). Limit kriging. Technometrics 48 458–466.
    Mathematical Reviews (MathSciNet): MR2328615
    Digital Object Identifier: doi:10.1198/004017006000000011
  • Joseph, V. R., Hung, Y. and Sudjianto, A. (2008). Blind kriging: A new method for developing metamodels. ASME Journal of Mechanical Design 130 031102–1–8.
  • Joseph, V. R. and Kang, L. (2011). Regression-based inverse distance weighting with applications to computer experiments. Technometrics 53 254–265.
    Mathematical Reviews (MathSciNet): MR2857703
    Digital Object Identifier: doi:10.1198/TECH.2011.09154
  • Paciorek, C. J. and Schervish, M. J. (2006). Spatial modelling using a new class of nonstationary covariance functions. Environmetrics 17 483–506.
    Mathematical Reviews (MathSciNet): MR2240939
    Digital Object Identifier: doi:10.1002/env.785
  • Peng, C. Y. and Wu, C. F. J. (2012). Regularized kriging. Unpublished manuscript.
  • Qian, P. Z. G., Seepersad, C. C., Joseph, V. R., Allen, J. K. and Wu, C. F. J. (2006). Building surrogate models with detailed and approximate simulations. ASME Journal of Mechanical Design 128 668–677.
  • Ranjan, P., Haynes, R. and Karsten, R. (2011). A computationally stable approach to Gaussian process interpolation of deterministic computer simulation data. Technometrics 53 366–378.
    Mathematical Reviews (MathSciNet): MR2850469
    Digital Object Identifier: doi:10.1198/TECH.2011.09141
  • Sacks, J., Welch, W. J., Mitchell, T. J. and Wynn, H. P. (1989). Design and analysis of computer experiments. Statist. Sci. 4 409–423.
    Mathematical Reviews (MathSciNet): MR1041765
    Digital Object Identifier: doi:10.1214/ss/1177012413
    Project Euclid: euclid.ss/1177012413
  • Sampson, P. D. and Guttorp, P. (1992). Nonparametric estimation of nonstationary spatial covariance structure. J. Amer. Statist. Assoc. 87 108–119.
  • Santner, T. J., Williams, B. J. and Notz, W. I. (2003). The Design and Analysis of Computer Experiments. Springer, New York.
    Mathematical Reviews (MathSciNet): MR2160708
    Zentralblatt MATH: 1041.62068
  • Schmidt, A. M. and O’Hagan, A. (2003). Bayesian inference for nonstationary spatial covariance structure via spatial deformations. J. R. Stat. Soc. Ser. B Stat. Methodol. 65 745–758.
    Mathematical Reviews (MathSciNet): MR1998632
    Zentralblatt MATH: 1063.62034
    Digital Object Identifier: doi:10.1111/1467-9868.00413
  • Wackernagel, H. (2003). Multivariate Geostatistics, 3rd ed. Springer, New York.
  • Welch, W. J., Buck, R. J., Sacks, J., Wynn, H. P., Mitchell, T. J. and Morris, M. D. (1992). Screening, predicting, and computer experiments. Technometrics 34 15–25.
  • Xiong, Y., Chen, W., Apley, D. W. and Ding, X. (2007). A non-stationary covariance-based kriging method for metamodelling in engineering design. Internat. J. Numer. Methods Engrg. 71 733–756.
  • Yamamoto, J. K. (2000). An alternative measure of the reliability of ordinary Kriging estimates. Math. Geol. 32 489–509.
    Mathematical Reviews (MathSciNet): MR1751576
    Zentralblatt MATH: 0971.86006
    Digital Object Identifier: doi:10.1023/A:1007577916868

The Institute of Mathematical Statistics

The Annals of Applied Statistics

New content alerts

Email RSS ToC RSS Article
  • You have access to this content.
  • You have partial access to this content.
  • You do not have access to this content.
Turn Off MathJax

What is MathJax?

More like this

+ See more