Fisher-Rao distance between truncated distributions and robustness analysis in uncertainty quantification

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ketema, Baalu Belay, Bousquet, Nicolas, Costantino, Francesco, Gamboa, Fabrice, Iooss, Bertrand, Sueur, Roman
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916341434286080
author Ketema, Baalu Belay
Bousquet, Nicolas
Costantino, Francesco
Gamboa, Fabrice
Iooss, Bertrand
Sueur, Roman
author_facet Ketema, Baalu Belay
Bousquet, Nicolas
Costantino, Francesco
Gamboa, Fabrice
Iooss, Bertrand
Sueur, Roman
contents Input variables in numerical models are often subject to several levels of uncertainty, usually modeled by probability distributions. In the context of uncertainty quantification applied to these models, studying the robustness of output quantities with respect to the input distributions requires: (a) defining variational classes for these distributions; (b) calculating boundary values for the output quantities of interest with respect to these variational classes. The latter should be defined in a way that is consistent with the information structure defined by the ``baseline'' choice of input distributions. Considering parametric families, the variational classes are defined using the geodesic distance in their Riemannian manifold, a generic approach to such problems. Theoretical results and application tools are provided to justify and facilitate the implementation of such robustness studies, in concrete situations where these distributions are truncated -- a setting frequently encountered in applications. The feasibility of our approach is illustrated in a simplified industrial case study.
format Preprint
id arxiv_https___arxiv_org_abs_2407_21542
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Fisher-Rao distance between truncated distributions and robustness analysis in uncertainty quantification
Ketema, Baalu Belay
Bousquet, Nicolas
Costantino, Francesco
Gamboa, Fabrice
Iooss, Bertrand
Sueur, Roman
Metric Geometry
Optimization and Control
Probability
Input variables in numerical models are often subject to several levels of uncertainty, usually modeled by probability distributions. In the context of uncertainty quantification applied to these models, studying the robustness of output quantities with respect to the input distributions requires: (a) defining variational classes for these distributions; (b) calculating boundary values for the output quantities of interest with respect to these variational classes. The latter should be defined in a way that is consistent with the information structure defined by the ``baseline'' choice of input distributions. Considering parametric families, the variational classes are defined using the geodesic distance in their Riemannian manifold, a generic approach to such problems. Theoretical results and application tools are provided to justify and facilitate the implementation of such robustness studies, in concrete situations where these distributions are truncated -- a setting frequently encountered in applications. The feasibility of our approach is illustrated in a simplified industrial case study.
title Fisher-Rao distance between truncated distributions and robustness analysis in uncertainty quantification
topic Metric Geometry
Optimization and Control
Probability
url https://arxiv.org/abs/2407.21542