Skip to Main content Skip to Navigation
Conference papers

First order Sobol indices for physical models via inverse regression

Abstract : In a bayesian inverse problem context, we aim at performing sensitivity analysis to help understand and adjust the physical model. To do so, we introduce indicators inspired by Sobol indices but focused on the inverse model. Since this inverse model is not generally available in closed form, we propose to use a parametric surrogate model to approximate it. The parameters of this model may be estimated via standard EM inference. Then we can exploit its tractable form and perform Monte-Carlo integration to efficiently estimate these pseudo Sobol indices.
Complete list of metadata

Cited literature [10 references]  Display  Hide  Download
Contributor : Benoit Kugler <>
Submitted on : Wednesday, September 30, 2020 - 3:35:56 PM
Last modification on : Thursday, March 25, 2021 - 2:13:40 PM
Long-term archiving on: : Thursday, December 31, 2020 - 6:01:14 PM


Files produced by the author(s)


  • HAL Id : hal-02951375, version 1



Benoit Kugler, Florence Forbes, Sylvain Douté. First order Sobol indices for physical models via inverse regression. JDS 2020 - 52èmes Journées de Statistique de la Société Française de Statistique (SFdS), Jun 2021, Nice, France. pp.1-6. ⟨hal-02951375⟩



Record views


Files downloads