English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

The FANOVA (or "Sobol'-Hoeffding") decomposition of multivariate functionshas been used for high-dimensional model representation and global sensitivityanalysis. When the objective function f has no simple analytic form and iscostly to evaluate, a practical limitation is that computing FANOVA terms maybe unaffordable due to numerical integration costs. Several approximateapproaches relying on random field models have been proposed to alleviate thesecosts, where f is substituted by a (kriging) predictor or by conditionalsimulations. In the present work, we focus on FANOVA decompositions of Gaussianrandom field sample paths, and we notably introduce an associated kerneldecomposition (into 2^{2d} terms) called KANOVA. An interpretation in terms oftensor product projections is obtained, and it is shown that projected kernelscontrol both the sparsity of Gaussian random field sample paths and thedependence structure between FANOVA effects. Applications on simulated datashow the relevance of the approach for designing new classes of covariancekernels dedicated to high-dimensional kriging.

On ANOVA decompositions of kernels and Gaussian random field paths