Skip to Main content Skip to Navigation
Conference papers

Estimation itérative en propagation d'incertitudes : réglage robuste de l'algorithme de Robbins-Monro

Abstract : In uncertainty quantification of numerical simulation models, the classical approach for quantile estimation requires availability of the full sample of the studied variable. This approach is not suitable at exascale as large ensembles of simulation runs would need to gather a prohibitively large amount of data. This problem can be solved thanks to an on-the-fly (iterative) approach based on the Robbins-Monro algorithm. We numerically study this algorithm for estimating a discretized quantile function from samples of limited size (a few hundreds observations). As in practice, the distribution of the underlying variable is unknown, the goal is to define "robust" values of the algorithm parameters, which means that quantile estimates have to be reasonably good in most situations.
Document type :
Conference papers
Complete list of metadata

Cited literature [9 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-02511787
Contributor : Bertrand Iooss <>
Submitted on : Thursday, March 19, 2020 - 9:40:47 AM
Last modification on : Saturday, November 7, 2020 - 9:47:46 AM
Long-term archiving on: : Saturday, June 20, 2020 - 1:08:59 PM

File

sfds2020_iooss.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02511787, version 1

Collections

INSMI | EDF | CNRS

Citation

Bertrand Iooss. Estimation itérative en propagation d'incertitudes : réglage robuste de l'algorithme de Robbins-Monro. 52èmes Journées de Statistiques de la Société Française de Statistique (SFdS), 2020, Nice, France. pp.466-471. ⟨hal-02511787⟩

Share

Metrics

Record views

62

Files downloads

68