×

Estimation of the derivative-based global sensitivity measures using a Gaussian process metamodel. (English) Zbl 1386.65059

Summary: Physical phenomena are often studied using numerical simulators. Such computer codes are a function of uncertain input parameters and a global sensitivity analysis (GSA) can be performed to identify their impacts on the simulator outputs. Sobol’ indices, based on output variance decomposition, are commonly used to perform quantitative GSA. For many years now, other tools have been studied, closer to physical practices such as the derivative-based global sensitivity measures (DGSM). However, numerical simulators rarely provide the output gradient and DGSM estimation is not directly possible. To address this limitation, we propose to estimate the DGSMs using a Gaussian process metamodel (GPM) which approximates the simulator. Based on this GPM, we propose two DGSM estimators: a plug-in one defined by the DGSM of the GPM predictor and another one defined by the expectation of the DGSM associated to the full-GPM. The latter is equal to the first one completed by a variance term and can be accompanied by a credibility interval. For Gaussian kernel and uniform input laws, analytical formulas are given for both DGSM estimators. For all other situations, Monte Carlo methods for the expectation approximations are proposed: a propagative version of the Gibbs sampler and a chi-square approximation. Moreover, a significance test for the full-GPM based estimator is proposed for screening. The convergence of the two GPM-based DGSM estimators and the Monte Carlo approaches is compared in analytical test cases. Finally, we apply our work to an environmental application.

MSC:

65C60 Computational problems in statistics (MSC2010)
62G99 Nonparametric inference
60G15 Gaussian processes
62P30 Applications of statistics in engineering and industry; control charts
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] R. H. Byrd, P. Lu, J. Nocedal, and C. Zhu, {\it A limited memory algorithm for bound constrained optimization}, SIAM J. Sci. Comput., 16 (1995), pp. 1190-1208. · Zbl 0836.65080
[2] F. Campolongo, J. Cariboni, and A. Saltelli, {\it An effective screening design for sensitivity analysis of large models}, Environ. Model. Softw., 22 (2007), pp. 1509-1518, http://dx.doi.org/10.1016/j.envsoft.2006.10.004, http://www.sciencedirect.com/science/article/pii/S1364815206002805.
[3] W. Chen, R. Jin, and A. Sudjianto, {\it Analytical metamodel-based global sensitivity analysis and uncertainty propagation for robust design}, J. Mech. Design, 127 (2005), pp. 875-886.
[4] P. G. Constantine, {\it Active Subspaces: Emerging Ideas for Dimension Reduction in Parameter Studies}, SIAM Spotlights, SIAM, Philadelphia, 2015. · Zbl 1431.65001
[5] S. Kucherenko and S. Song, {\it Derivative-based global sensitivity measures and their link with Sobol’ sensitivity indices}, in Proceedings of the 11th International Conference on Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing (MCQMC 2014), Leuven, Belgium, 2015. · Zbl 1356.65010
[6] S. Da Veiga, {\it Global sensitivity analysis with dependence measures}, J. Stat. Comput. Simul., 85 (2015), pp. 1283-1305. · Zbl 1457.62188
[7] M. De Lozzo and A. Marrel, {\it New improvements in the use of dependence measures for sensitivity analysis and screening}, J. Stat. Comput. Simul., (2016). · Zbl 1386.65059
[8] E. de Rocquigny, N. Devictor, and S. Tarantola, {\it Uncertainty in Industrial Practice}, Wiley, New York, 2008. · Zbl 1161.90001
[9] G. Dreyfus, {\it Neural Networks— Methodology and Applications}, Springer, New York, 2005. · Zbl 1119.92003
[10] B. Efron and C. Stein, {\it The jackknife estimate of variance}, Ann. Statist., 9 (1981), pp. 586-596, http://dx.doi.org/10.1214/aos/1176345462. · Zbl 0481.62035
[11] K.-T. Fang, R. Li, and A. Sudjianto, {\it Design and Modeling for Computer Experiments}, Computer Science and Data Analysis Series, Chapman & Hall/CRC, Boca Raton, FL, 2006. · Zbl 1093.62117
[12] A. Forrester, A. Sobester, and A. Keane, {\it Engineering Design via Surrogate Modelling: A Practical Guide}, Wiley, New York, 2008.
[13] J.-C. Fort, T. Klein, and N. Rachdi, {\it New sensitivity analysis subordinated to a contrast}, Comm. Statist. Theory Methods, (2015). · Zbl 1397.62592
[14] S. Geman and D. Geman, {\it Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images}, in Proceedings of the IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI-6, 1984, pp. 721-741, http://dx.doi.org/10.1109/TPAMI.1984.4767596. · Zbl 0573.62030
[15] G. Gretton, O. Bousquet, A. Smola, and B. Schölkopf, {\it Measuring statistical dependence with Hilbert-Schmidt norms}, in Proceedings of Algorithmic Learning Theory, Springer-Verlag, 2005, pp. 63-77. · Zbl 1168.62354
[16] T. Hastie, R. Tibshirani, and J. Friedman, {\it The Elements of Statistical Learning}, Springer Ser. Statist., Springer, New York, 2001. · Zbl 0973.62007
[17] T. Homma and A. Saltelli, {\it Importance measures in global sensitivity analysis of nonlinear models}, Reliability Engineering System Safety, 52 (1996), pp. 1-17, http://dx.doi.org/10.1016/0951-8320(96)00002-6, http://www.sciencedirect.com/science/article/pii/0951832096000026.
[18] B. Iooss and P. Lemaître, {\it A review on global sensitivity analysis methods}, in Uncertainty Management in Simulation-Optimization of Complex Systems: Algorithms and Applications, C. Meloni and G. Dellino, eds., New York, Springer, 2015.
[19] B. Iooss, A.-L. Popelin, G. Blatman, C. Ciric, F. Gamboa, S. Lacaze, and M. lamboni, {\it Some new insights in derivative-based global sensitivity measures}, in Proceedings of SPAM11 & ESREL 2012 Conference, ESRA, 2012.
[20] A. Janon, M. Nodet, and C. Prieur, {\it Uncertainties assessment in global sensitivity indices estimation from metamodels}, Int. J. Uncertain. Quantif., 4 (2014). · Zbl 1357.65007
[21] R. Jin, W. Chen, and A. Agus, {\it An efficient algorithm for constructing optimal design of computer experiments}, J. Statist. Planning Inference, 134 (2005), pp. 268-287. · Zbl 1066.62075
[22] A. Kiparissides, S. S. Kucherenko, A. Mantalaris, and E. N. Pistikopoulos, {\it Global sensitivity analysis challenges in biological systems modeling}, Industrial Engineering Chemistry Research, 48 (2009), pp. 7168-7180, http://dx.doi.org/10.1021/ie900139x.
[23] J. P. Kleijnen, {\it Sensitivity analysis and related analyses: A review of some statistical techniques}, J. Statist. Comput. Simul., 57 (1997), pp. 111-142. · Zbl 1102.62314
[24] S. Kucherenko, M. Rodriguez-Fernandez, C. Pantelides, and N. Shah, {\it Monte Carlo evaluation of derivative-based global sensitivity measures}, Reliability Engineering and System Safety, 94 (2009), pp. 1135 – 1148, http://dx.doi.org/10.1016/j.ress.2008.05.006, http://www.sciencedirect.com/science/article/pii/S095183200800152X.
[25] M. Lamboni, B. Iooss, A.-L. Popelin, and F. Gamboa, {\it Derivative-based global sensitivity measures: General links with Sobol’ indices and numerical tests}, Math. Comput. Simul., 87 (2013), pp. 45-54. · Zbl 1490.62035
[26] L. Lantuéjoul and N. Desassis, {\it Simulation of a Gaussian random vector: A propagative version of the Gibbs sampler}, in Proceedings of the Ninth International Geostatistics Congress, Oslo, Norway, 2012.
[27] L. Le Gratiet, C. Cannamela, and B. Iooss, {\it A Bayesian approach for global sensitivity analysis of (multifidelity) computer codes}, SIAM/ASA J. Uncertainty Quantification, 2 (2014), pp. 336-363. · Zbl 1308.62107
[28] H. Liu, Y. Tang, and H. H. Zhang, {\it A new chi-square approximation to the distribution of non-negative definite quadratic forms in non-central normal variables}, Comput. Statist. Data Anal., 53 (2009), pp. 853 – 856, http://dx.doi.org/10.1016/j.csda.2008.11.025, http://www.sciencedirect.com/science/article/pii/S0167947308005653. · Zbl 1364.62041
[29] A. Marrel, B. Iooss, B. Laurent, and O. Roustant, {\it Calculations of Sobol indices for the Gaussian process metamodel}, Reliability Engineering System Safety, 94 (2009), pp. 742-751, http://dblp.uni-trier.de/db/journals/ress/ress94.html#MarrelILR09.
[30] A. Marrel, B. Iooss, F. Van Dorpe, and E. Volkova, {\it An efficient methodology for modeling complex computer codes with Gaussian processes}, Comput. Statist. Data Anal., 52 (2008), pp. 4731-4744, http://dx.doi.org/10.1016/j.csda.2008.03.026, https://hal.archives-ouvertes.fr/hal-00239492. · Zbl 1452.62094
[31] M. D. McKay, R. J. Beckman, and W. J. Conover, {\it A comparison of three methods for selecting values of input variables in the analysis of output from a computer code}, Technometrics, 21 (1979), pp. 239-245, http://dx.doi.org/10.2307/1268522. · Zbl 0415.62011
[32] M. D. Morris, {\it Factorial sampling plans for preliminary computational experiments}, Technometrics, 33 (1991), pp. 161-174, http://dx.doi.org/10.2307/1269043.
[33] M. D. Morris, L. M. Moore, and M. D. McKay, {\it Sampling plans based on balanced incomplete block designs for evaluating the importance of computer model inputs}, J. Statist. Planning Inference, 136 (2006), pp. 3203-3220, http://www.sciencedirect.com/science/article/B6V0M-4FGX4CC-1/1/361af9baf3feb5e5e7c02c9f7a534d89. · Zbl 1094.62087
[34] J. Oakley and A. O’Hagan, {\it Probabilistic sensitivity analysis of complex models: A Bayesian approach}, J. Roy. Statist. Soc., Ser. B, 66 (2004), pp. 751-769. · Zbl 1046.62027
[35] E. Plischke, E. Borgonovo, and C. L. Smith, {\it Global sensitivity measures from given data}, European J. Oper. Res., 226 (2013), pp. 536-550, http://dx.doi.org/http://dx.doi.org/10.1016/j.ejor.2012.11.047, http://www.sciencedirect.com/science/article/pii/S0377221712008995. · Zbl 1292.90290
[36] C. E. Rasmussen and C. K. I. Williams, {\it Gaussian Processes for Machine Learning (Adaptive Computation and Machine Learning)}, MIT Press, Cambridge, MA, 2005.
[37] O. Roustant, J. Fruth, B. Iooss, and S. Kuhnt, {\it Crossed-derivative based sensitivity measures for interaction screening}, Math. Comput. Simul., 105 (2014), pp. 105-118, http://dx.doi.org/10.1016/j.matcom.2014.05.005, http://www.sciencedirect.com/science/article/pii/S0378475414001244. · Zbl 1278.62109
[38] R. Roustant, D. Ginsbourger, and Y. Deville, {\it DiceKriging, DiceOptim: Two R packages for the analysis of computer experiments by Kriging-based metamodeling and optimization}, J. Statist. Softw., 51 (2012), pp. 1-55, http://www.jstatsoft.org/v51/i01/.
[39] A. Saltelli, K. Chan, and E. M. Scott, {\it Sensitivity Analysis}, Wiley, New York, 2000. · Zbl 0961.62091
[40] A. Saltelli, M. Ratto, T. Andres, F. Campolongo, J. Cariboni, D. Gatelli, M. Saisana, and S. Tarantola, {\it Global Sensitivity Analysis: The Primer}, Wiley, New York, 2008. · Zbl 1161.00304
[41] A. Saltelli, S. Tarantola, F. Campolongo, and M. Ratto, {\it Sensitivity Analysis in Practice: A Guide to Assessing Scientific Models}, Wiley, New York, 2004. · Zbl 1049.62112
[42] T. J. Santner, B. J. Williams, and W. Notz, {\it The Design and Analysis of Computer Experiments}, Springer Ser. Statist., Springer, New York, 2003. · Zbl 1041.62068
[43] I. Sobol and A. Gersham, {\it On an alternative global sensitivity estimator}, in Proceedings of SAMO 1995, Belgirate, Italy, 1995, pp. 40-42.
[44] I. Sobol and S. Kucherenko, {\it Derivative based global sensitivity measures and their link with global sensitivity indices}, Math. Comput. Simul., 79 (2009), pp. 3009-3017, http://dx.doi.org/10.1016/j.matcom.2009.01.023, http://www.sciencedirect.com/science/article/pii/S0378475409000354. · Zbl 1167.62005
[45] I. M. Sobol, {\it Sensitivity estimates for nonlinear mathematical models}, Math. Model. Civil Engrg., 1 (1993), pp. 407-414. · Zbl 1039.65505
[46] C. Soize and R. Ghanem, {\it Physical systems with random uncertainties: Chaos representations with arbitrary probability measure}, SIAM J. Sci. Comput., 26 (2004), pp. 395-410, http://dx.doi.org/10.1137/S1064827503424505. · Zbl 1075.60084
[47] B. Sudret, {\it Global sensitivity analysis using polynomial chaos expansions}, Reliability Engineering System Safety, 93 (2008), pp. 964-979, http://dx.doi.org/10.1016/j.ress.2007.04.002, http://www.sciencedirect.com/science/article/pii/S0951832007001329.
[48] B. Sudret and C. V. Mai, {\it Computing derivative-based global sensitivity measures using polynomial chaos expansions}, Reliability Engineering System Safety, 134 (2015), pp. 241-250.
[49] G. J. Székely, M. L. Rizzo, and N. K. Bakirov, {\it Measuring and testing dependence by correlation of distances}, Ann. Statist., 35 (2007), pp. 2769-2794. · Zbl 1129.62059
[50] S. Touzani and D. Busby, {\it Screening method using the derivative-based global sensitivity indices with application to reservoir simulator}, Oil Gas Sci. Technol., 69 (2014), pp. 619-632.
[51] E. Volkova, B. Iooss, and F. Van Dorpe, {\it Global sensitivity analysis for a numerical model of radionuclide migration from the RRC Kurchatov Institute radwaste disposal site}, Stoch. Environ. Research Risk Assessment, 22 (2008), pp. 17-31, http://dx.doi.org/10.1007/s00477-006-0093-y. · Zbl 1169.62399
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.