The Gaussian Process Modeling Module in UQLab

Document Type : Regular Article

Authors

1 Chair of Risk, Safety and Uncertainty Quantification, ETH Zurich, Stefano-Franscini-Platz 5, 8093 Zurich, Switzerland

2 ETH Zurich, Stefano-Franscini-Platz 5, 8093 Zurich, Switzerland

Abstract

We introduce the Gaussian process (GP) modeling module developed within the UQLab software framework. The novel design of the GP-module aims at providing seamless integration of GP modeling into any uncertainty quantification workflow, as well as a standalone surrogate modeling tool. We first briefly present the key mathematical tools on the basis of GP modeling (a.k.a. Kriging), as well as the associated theoretical and computational framework. We then provide an extensive overview of the available features of the software and demonstrate its flexibility and user-friendliness. Finally, we showcase the usage and the performance of the software on several applications borrowed from different fields of engineering. These include a basic surrogate of a well-known analytical benchmark function; a hierarchical Kriging example applied to wind turbine aero-servo-elastic simulations and a more complex geotechnical example that requires a non-stationary, user-defined correlation function. The GP-module, like the rest of the scientific code that is shipped with UQLab, is open source (BSD license).

Highlights

Google Scholar

Keywords

Main Subjects

References

[1]     Lemaire M, Chateauneuf A, Mitteau J-C. Structural Reliability. London, UK: ISTE; 2009. doi:10.1002/9780470611708.
[2]     Saltelli A, Chan K, Scott EM. Sensitivity Analysis. J. Wiley & Sons; 2000.
[3]     Tsompanakis Y, Lagaros ND, Papadrakakis M. Structural Design Optimization Considering Uncertainties. CRC Press; 2008.
[4]     Dashti M, Stuart AM. The Bayesian Approach to Inverse Problems. Handb. Uncertain. Quantif., Cham: Springer International Publishing; 2017, p. 311–428. doi:10.1007/978-3-319-12385-1_7.
[5]     Santner TJ, Williams BJ, Notz WI. The Design and Analysis of Computer Experiments. New York, NY: Springer New York; 2003. doi:10.1007/978-1-4757-3799-8.
[6]     Fang K-T, Li R, Sudjianto A. Design and Modeling for Computer Experiments. Chapman and Hall/CRC; 2005. doi:10.1201/9781420034899.
[7]     Forrester AIJ, Sbester A, Keane AJ. Engineering Design via Surrogate Modelling. Chichester, UK: John Wiley & Sons, Ltd; 2008. doi:10.1002/9780470770801.
[8]     Sacks J, Welch WJ, Mitchell TJ, Wynn HP. Design and analysis of computer experiments. Stat Sci 1989;4:409–23.
[9]     Ghanem RG, Spanos PD. Stochastic Finite Elements: A Spectral Approach. New York, NY: Springer New York; 1991. doi:10.1007/978-1-4612-3094-6.
[10]    Xiu D, Karniadakis GE. The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations. SIAM J Sci Comput 2002;24:619–44. doi:10.1137/S1064827501387826.
[11]    Vapnik VN. The Nature of Statistical Learning Theory. New York, NY: Springer New York; 2000. doi:10.1007/978-1-4757-3264-1.
[12]    Krige DG. A statistical approach to some basic mine valuation problems on the Witwatersrand. J South African Inst Min Metall 1951;52:119–39.
[13]    Welch WJ, Buck RJ, Sacks J, Wynn HP, Mitchell TJ, Morris MD. Screening, Predicting, and Computer Experiments. Technometrics 1992;34:15–25. doi:10.1080/00401706.1992.10485229.
[14]    Marrel A, Iooss B, Van Dorpe F, Volkova E. An efficient methodology for modeling complex computer codes with Gaussian processes. Comput Stat Data Anal 2008;52:4731–44. doi:10.1016/j.csda.2008.03.026.
[15]    Gaspar B, Teixeira AP, Soares CG. Assessment of the efficiency of Kriging surrogate models for structural reliability analysis. Probabilistic Eng Mech 2014;37:24–34. doi:10.1016/j.probengmech.2014.03.011.
[16]    Iooss B, Lemaître P. A Review on Global Sensitivity Analysis Methods, 2015, p. 101–22. doi:10.1007/978-1-4899-7547-8_5.
[17]    Gratiet L Le, Marelli S, Sudret B. Metamodel-Based Sensitivity Analysis: Polynomial Chaos Expansions and Gaussian Processes. Handb. Uncertain. Quantif., Cham: Springer International Publishing; 2015, p. 1–37. doi:10.1007/978-3-319-11259-6_38-1.
[18]    Moustapha M, Bourinet J-M, Guillaume B, Sudret B. Comparative Study of Kriging and Support Vector Regression for Structural Engineering Applications. ASCE-ASME J Risk Uncertain Eng Syst Part A Civ Eng 2018;4:04018005. doi:10.1061/AJRUA6.0000950.
[19]    Echard B, Gayton N, Lemaire M. AK-MCS: An active learning reliability method combining Kriging and Monte Carlo Simulation. Struct Saf 2011;33:145–54. doi:10.1016/j.strusafe.2011.01.002.
[20]    Dubourg V, Sudret B. Meta-model-based importance sampling for reliability sensitivity analysis. Struct Saf 2014;49:27–36. doi:10.1016/j.strusafe.2013.08.010.
[21]    Simpson TW, Mauery TM, Korte JJ, Mistree F. Kriging Models for Global Approximation in Simulation-Based Multidisciplinary Design Optimization. AIAA J 2001;39:2233–41. doi:10.2514/2.1234.
[22]    Moustapha M, Sudret B, Bourinet J-M, Guillaume B. Quantile-based optimization under uncertainties using adaptive Kriging surrogate models. Struct Multidiscip Optim 2016;54:1403–21. doi:10.1007/s00158-016-1504-4.
[23]    Bachoc F, Bois G, Garnier J, Martinez J-M. Calibration and Improved Prediction of Computer Models by Universal Kriging. Nucl Sci Eng 2014;176:81–97. doi:10.13182/NSE12-55.
[24]    Rasmussen CE, Williams CK. Gaussian processes for machine learning (adaptive computation and machine learning) Cambridge, Cambridge, MA, USA: MIT Press 2005.
[25]    Deutsch C V, Journel AG. GSLIB: Geostatistical SoftwareLibrary and User’s Guide. New York, US: Oxford University Press; 1992.
[26]    Roustant O, Ginsbourger D, Deville Y. Dicekriging, Diceoptim: Two R packages for the analysis of computer experiments by kriging-based metamodelling and optimization. J Stat Softw 2012;51:1–55.
[27]    Dupuy D, Helbert C, Franco J. DiceDesign and DiceEval: Two R packages for design and analysis of computer experiments. J Stat Softw 2015;65:1–38.
[28]    Picheny V, Ginsbourger D, Roustant O. DiceOptim: Kriging-Based Optimization for Computer Experiments. R package version 0.8-1 2016.
[29]    Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O, et al. Scikit-learn: Machine learning in Python. J Mach Learn Res 2011;12:2825–30.
[30]    Paulson C, Ragkousis G. pyKriging: A Python Kriging Toolkit [Data set]. 2015.
[31]    GPy: A gaussian process framework in python 2012.
[32]    Lophaven SN, Nielsen HB, Sondergaard J. Aspects of the matlab toolbox DACE. Technical report, Informatics and Mathematical Modelling. 2002.
[33]    Couckuyt I, Dhaene T, Demeester P. ooDACE toolbox: a flexible object-oriented Kriging implementation. J Mach Learn Res 2014;15:3183–6.
[34]    Bect J, Vazquez E. STK: a Small (Matlab/Octave) Toolbox for Kriging 2014.
[35]    Rasmussen CE, Nickisch H. Gaussian processes for machine learning (GPML) toolbox. J Mach Learn Res 2010;11:3011–5.
[36]    Marelli S, Sudret B. UQLab: A Framework for Uncertainty Quantification in Matlab. Vulnerability, Uncertainty, Risk (Proc. 2nd Int. Conf. Vulnerability, Risk Anal. Manag. (ICVRAM2014), Liverpool, United Kingdom), Reston, VA: American Society of Civil Engineers; 2014, p. 2554–63. doi:10.1061/9780784413609.257.
[37]    Schöbi R, Sudret B, Wiart J. Polynomial-chaos-based Kriging. Int J Uncertain Quantif 2015;5:171–93.
[38]    Stein ML. Interpolation of Spatial Data. New York, NY: Springer New York; 1999. doi:10.1007/978-1-4612-1494-6.
[39]    Bachoc F. Cross Validation and Maximum Likelihood estimations of hyper-parameters of Gaussian processes with model misspecification. Comput Stat Data Anal 2013;66:55–69. doi:10.1016/j.csda.2013.03.016.
[40]    Dubrule O. Cross validation of kriging in a unique neighborhood. J Int Assoc Math Geol 1983;15:687–99. doi:10.1007/BF01033232.
[41]    Cressie NAC. Statistics for Spatial Data. Hoboken, NJ, USA: John Wiley & Sons, Inc.; 1993. doi:10.1002/9781119115151.
[42]    Sudret B. Uncertainty propagation and sensitivity analysis in mechanical models–Contributions to structural reliability and stochastic spectral methods. 2007.
[43]    De Rocquigny E, Devictor N, Tarantola S. Uncertainty in industrial practice: a guide to quantitative uncertainty management. John Wiley & Sons; 2008.
[44]    Lataniotis C, Marelli S, Sudret B. UQLab user manual–Kriging (Gaussian process modelling). 2017.
[45]    Marelli S, Schöbi R, Sudret B. Uqlab user manual - structural reliability. Technical report, Chair of Risk, Safety and Uncertainty Quantification, ETHZurich. Report UQLab-V0.92-107; 2017.
[46]    Jones DR, Schonlau M, Welch WJ. Efficient Global Optimization of Expensive Black-Box Functions. J Glob Optim 1998;13:455–92. doi:10.1023/A:1008306431147.
[47]    McKay MD, Beckman RJ, Conover WJ. Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code. Technometrics 1979;21:239–45. doi:10.1080/00401706.1979.10489755.
[48]    Han Z, Zimmerman R, Görtz S. Alternative Cokriging Method for Variable-Fidelity Surrogate Modeling. AIAA J 2012;50:1205–10. doi:10.2514/1.J051243.
[49]    Abdallah I, Sudret B, Lataniotis C, Sørensen JD, Natarajan A. Fusing simulation results from multifidelity aero-servo-elastic simulators-Application to extreme loads on wind turbine. Proc. 12th Int. Conf. Appl. Stat. Probab. Civ. Eng. (ICASP12), Vancouver, Canada, July 12-15, University of British Columbia; 2015.
[50]    Schöbi R, Marelli S, Sudret B. Uqlab user manual–pc-kriging, Technical report, Chair of Risk, Safety & Uncertainty Quantification, ETH Zurich. Report UQLab-V1. 0109. Technical report, Chair of Risk, Safety & Uncertainty Quantification, ETH Zurich; 2017.

Files

Cited by

History

  • Receive Date: 04 May 2018
  • Revise Date: 20 June 2018
  • Accept Date: 20 June 2018

Share

How to cite

Statistics