Risk of estimators for Sobol' sensitivity indices based on metamodels

被引:2
|
作者
Panin, Ivan [1 ,2 ]
机构
[1] Skolkovo Inst Sci & Technol, Bolshoy Blvd 30,Bld 1, Moscow 121205, Russia
[2] Kharkevich Inst Informat Transmiss Problems, Bolshoy Karetny 19,Bld 1, Moscow 127051, Russia
来源
ELECTRONIC JOURNAL OF STATISTICS | 2021年 / 15卷 / 01期
关键词
Global sensitivity analysis; Sobol' indices; polynomial chaos approximation; OPTIMAL GLOBAL RATES; CONVERGENCE; REGRESSION; MODELS;
D O I
10.1214/20-EJS1793
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Sobol' sensitivity indices allow to quantify the respective effects of random input variables and their combinations on the variance of mathematical model output. We focus on the problem of Sobol' indices estimation via a metamodeling approach where we replace the true mathematical model with a sample-based approximation to compute sensitivity indices. We propose a new method for indices quality control and obtain asymptotic and non-asymptotic risk bounds for Sobol' indices estimates based on a general class of metamodels. Our analysis is closely connected with the problem of nonparametric function fitting using the orthogonal system of functions in the random design setting. It considers the relation between the metamodel quality and the error of the corresponding estimator for Sobol' indices and shows the possibility of fast convergence rates in the case of noiseless observations. The theoretical results are complemented with numerical experiments for the approximations based on multivariate Legendre and Trigonometric polynomials.
引用
收藏
页码:235 / 281
页数:47
相关论文
共 50 条
  • [31] Higher order Sobol' indices
    Owen, Art B.
    Dick, Josef
    Chen, Su
    INFORMATION AND INFERENCE-A JOURNAL OF THE IMA, 2014, 3 (01) : 59 - 81
  • [32] Sensitivity of the hygrothermal behaviour of homogeneous masonry constructions: from Sobol indices to decision trees
    Calle, Klaas
    Van den Bossche, Nathan
    12TH NORDIC SYMPOSIUM ON BUILDING PHYSICS (NSB 2020), 2020, 172
  • [33] Constrained Global Sensitivity Analysis: Sobol' indices for problems in non-rectangular domains
    Klymenko, Oleksiy V.
    Kucherenko, Sergei
    Shah, Nilay
    27TH EUROPEAN SYMPOSIUM ON COMPUTER AIDED PROCESS ENGINEERING, PT A, 2017, 40A : 151 - 156
  • [34] Sensitivity analysis using Sobol indices for the thermal modelling of an electrical machine for sizing by optimization
    Le Guyadec, Mathias
    Gerbaud, Laurent
    Vinot, Emmanuel
    Delinchant, Benoit
    COMPEL-THE INTERNATIONAL JOURNAL FOR COMPUTATION AND MATHEMATICS IN ELECTRICAL AND ELECTRONIC ENGINEERING, 2019, 38 (03) : 965 - 976
  • [35] A note on the Sobol' indices and interactive criteria
    Grabisch, Michel
    Labreuche, Christophe
    FUZZY SETS AND SYSTEMS, 2017, 315 : 99 - 108
  • [36] Reliable error estimation for Sobol’ indices
    Lluís Antoni Jiménez Rugama
    Laurent Gilquin
    Statistics and Computing, 2018, 28 : 725 - 738
  • [37] Different numerical estimators for main effect global sensitivity indices
    Kucherenko, S.
    Song, S.
    RELIABILITY ENGINEERING & SYSTEM SAFETY, 2017, 165 : 222 - 238
  • [38] Enhancing Stochastic Kriging Metamodels with Gradient Estimators
    Chen, Xi
    Ankenman, Bruce E.
    Nelson, Barry L.
    OPERATIONS RESEARCH, 2013, 61 (02) : 512 - 528
  • [39] Reliable error estimation for Sobol' indices
    Rugama, Lluis Antoni Jimenez
    Gilquin, Laurent
    STATISTICS AND COMPUTING, 2018, 28 (04) : 725 - 738
  • [40] Variance Components and Generalized Sobol' Indices
    Owen, Art B.
    SIAM-ASA JOURNAL ON UNCERTAINTY QUANTIFICATION, 2013, 1 (01): : 19 - 41