Non linear methods for inverse statistical problems

• Published in: Computational statistics & data analysis
• Main Authors: Barbillon, Pierre, Celeux, Gilles, Grimaud, Agnès, Lefebvre, Yannick, de Rocquigny, Etienne
• Format: Journal Article
• Language: English
• Published: Elsevier 2011
• Subjects: Mathematics; Statistics; Statistics Theory; Uncertainty Modelling; Non linear Approximation; Stochastic Algorithm; Kriging
• ISSN: 0167-9473
• DOI: 10.1016/j.csda.2010.05.030

In the uncertainty treatment framework considered in this paper, the intrinsic variability of the inputs of a physical simulation model is modelled by a multivariate probability distribution. The objective is to identify this probability distribution - the dispersion of which is independent of the sample size since intrinsic variability is at stake - based on observation of some model outputs. Moreover, in order to limit to a reasonable level the number of (usually burdensome) physical model runs inside the inversion algorithm, a non linear approximation methodology making use of Kriging and stochastic EM algorithm is presented. It is compared with iterated linear approximation on the basis of numerical experiments on simulated data sets coming from a simplified but realistic modelling of a dyke overflow. Situations where this non linear approach is to be preferred to linearisation are highlighted.