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
  • [41] ROBUSTNESS OF THE SOBOL' INDICES TO DISTRIBUTIONAL UNCERTAINTY
    Hart, Joseph
    Gremaud, Pierre
    INTERNATIONAL JOURNAL FOR UNCERTAINTY QUANTIFICATION, 2019, 9 (05) : 453 - 469
  • [42] Efficient Global Sensitivity Analysis of Model-Based Ultrasonic Nondestructive Testing Systems Using Machine Learning and Sobol' Indices
    Nagawkar, Jethro
    Leifsson, Leifur
    JOURNAL OF NONDESTRUCTIVE EVALUATION, DIAGNOSTICS AND PROGNOSTICS OF ENGINEERING SYSTEMS, 2021, 4 (04):
  • [43] SHAPLEY EFFECTS FOR SENSITIVITY ANALYSIS WITH CORRELATED INPUTS: COMPARISONS WITH SOBOL' INDICES, NUMERICAL ESTIMATION AND APPLICATIONS
    Iooss, Bertrand
    Prieur, Clementine
    INTERNATIONAL JOURNAL FOR UNCERTAINTY QUANTIFICATION, 2019, 9 (05) : 493 - 514
  • [44] Adaptive Design of Experiments for Sobol Indices Estimation Based on Quadratic Metamodel
    Burnaev, Evgeny
    Panin, Ivan
    STATISTICAL LEARNING AND DATA SCIENCES, 2015, 9047 : 86 - 95
  • [45] Structural integrity assessment of CANDU pressure tubes using Sobol indices for global sensitivity analysis
    Pelayo, R. E. Ortega
    El Bouzidi, S.
    Trottier, A.
    Forgues, F.
    Pfeiffer, S.
    NUCLEAR ENGINEERING AND DESIGN, 2024, 428
  • [46] Sensitivity Analysis for Epistemic Uncertain System Based on Sobol' Method
    锁斌
    郭惠昕
    Journal of Donghua University(English Edition), 2018, (04) : 344 - 347
  • [47] ASYMPTOTIC NORMALITY AND EFFICIENCY OF TWO SOBOL INDEX ESTIMATORS
    Janon, Alexandre
    Klein, Thierry
    Lagnoux, Agnes
    Nodet, Maelle
    Prieur, Clementine
    ESAIM-PROBABILITY AND STATISTICS, 2014, 18 : 342 - 364
  • [48] Efficient estimators for likelihood ratio sensitivity indices of complex stochastic dynamics
    Arampatzis, Georgios
    Katsoulakis, Markos A.
    Rey-Bellet, Luc
    JOURNAL OF CHEMICAL PHYSICS, 2016, 144 (10):
  • [49] COMPUTATION OF SOBOL INDICES IN GLOBAL SENSITIVITY ANALYSIS FROM SMALL DATA SETS BY PROBABILISTIC LEARNING ON MANIFOLDS
    Arnst, Maarten
    Soize, Christian
    Bulthuis, Kevin
    INTERNATIONAL JOURNAL FOR UNCERTAINTY QUANTIFICATION, 2021, 11 (02) : 1 - 23
  • [50] Computing Sobol indices in probabilistic graphical models
    Ballester-Ripoll, Rafael
    Leonelli, Manuele
    RELIABILITY ENGINEERING & SYSTEM SAFETY, 2022, 225