Sensitivity analysis for multidimensional and functional outputs
Applications
2013-11-15 v2 Statistics Theory
Methodology
Statistics Theory
Abstract
Let be random objects (the inputs), defined on some probability space and valued in some measurable space . Further, let be the output. Here, is a measurable function from to some Hilbert space ( could be either of finite or infinite dimension). In this work, we give a natural generalization of the Sobol indices (that are classically defined when ), when the output belongs to . These indices have very nice properties. First, they are invariant. under isometry and scaling. Further they can be, as in dimension , easily estimated by using the so-called Pick and Freeze method. We investigate the asymptotic behaviour of such estimation scheme.
Cite
@article{arxiv.1311.1797,
title = {Sensitivity analysis for multidimensional and functional outputs},
author = {Fabrice Gamboa and Alexandre Janon and Thierry Klein and Agnès Lagnoux},
journal= {arXiv preprint arXiv:1311.1797},
year = {2013}
}
Comments
Fixed missing references