Geodesic
Metamodel construction for sensitivity analysis
Sylvie Huet1 ; Marie-Luce Taupin2, 1
1UR 1404 MaIAGE, INRA Jouy-en-Josas, France; e-mail:
2Laboratoire LaMME, UMR CNRS 8071- USC INRA, Université d’Evry Val d’Essonne, France; e-mail:
ESAIM. Proceedings, Tome 60 (2017), pp. 27-69
Cet article a éte moissonné depuis la source EDP Sciences

Voir la notice de l'article

We propose to estimate a metamodel and the sensitivity indices of a complex model m in the Gaussian regression framework. Our approach combines methods for sensitivity analysis of complex models and statistical tools for sparse non-parametric estimation in multivariate Gaussian regression model. It rests on the construction of a metamodel for aproximating the Hoeffding-Sobol decomposition of m. This metamodel belongs to a reproducing kernel Hilbert space constructed as a direct sum of Hilbert spaces leading to a functional ANOVA decomposition. The estimation of the metamodel is carried out via a penalized least-squares minimization allowing to select the subsets of variables that contribute to predict the output. It allows to estimate the sensitivity indices of m. We establish an oracle-type inequality for the risk of the estimator, describe the procedure for estimating the metamodel and the sensitivity indices, and assess the performances of the procedure via a simulation study.
DOI : 10.1051/proc/201760027

Sylvie Huet  1   ; Marie-Luce Taupin  2 , 1

1 UR 1404 MaIAGE, INRA Jouy-en-Josas, France; e-mail:
2 Laboratoire LaMME, UMR CNRS 8071- USC INRA, Université d’Evry Val d’Essonne, France; e-mail: 
@article{EP_2017_60_a2,
     author = {Sylvie Huet and Marie-Luce Taupin},
     title = {Metamodel construction for sensitivity analysis},
     journal = {ESAIM. Proceedings},
     pages = {27--69},
     year = {2017},
     volume = {60},
     doi = {10.1051/proc/201760027},
     language = {en},
     url = {http://geodesic.mathdoc.fr/articles/10.1051/proc/201760027/}
}
TY  - JOUR
AU  - Sylvie Huet
AU  - Marie-Luce Taupin
TI  - Metamodel construction for sensitivity analysis
JO  - ESAIM. Proceedings
PY  - 2017
SP  - 27
EP  - 69
VL  - 60
UR  - http://geodesic.mathdoc.fr/articles/10.1051/proc/201760027/
DO  - 10.1051/proc/201760027
LA  - en
ID  - EP_2017_60_a2
ER  - 
%0 Journal Article
%A Sylvie Huet
%A Marie-Luce Taupin
%T Metamodel construction for sensitivity analysis
%J ESAIM. Proceedings
%D 2017
%P 27-69
%V 60
%U http://geodesic.mathdoc.fr/articles/10.1051/proc/201760027/
%R 10.1051/proc/201760027
%G en
%F EP_2017_60_a2
Sylvie Huet; Marie-Luce Taupin. Metamodel construction for sensitivity analysis. ESAIM. Proceedings, Tome 60 (2017), pp. 27-69. doi: 10.1051/proc/201760027

Cité par Sources :