ANOVA kernels and RKHS of zero mean functions for model-based sensitivity analysis

DOI10.1016/J.JMVA.2012.08.016MaRDI QIDQ1941425zbMATH OpenOpenAlexFDO

Authors Nicolas Durrande, D. Ginsbourger, O. Roustant, Laurent Carraro

Publication date 12 March 2013

Published in Journal of Multivariate Analysis (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1106.3571

zbMATH Keywords

Gaussian process regressionglobal sensitivity analysisHoeffdin-Sobol decompositionSS-ANOVA

Mathematics Subject Classification ID

Linear regression; mixed models (62J05) Gaussian processes (60G15) Analysis of variance and covariance (ANOVA) (62J10) Applications of functional analysis in probability theory and statistics (46N30) Sensitivity analysis for optimization problems on manifolds (49Q12)

Abstract: Given a reproducing kernel Hilbert space H of real-valued functions and a suitable measure mu over the source space D (subset of R), we decompose H as the sum of a subspace of centered functions for mu and its orthogonal in H. This decomposition leads to a special case of ANOVA kernels, for which the functional ANOVA representation of the best predictor can be elegantly derived, either in an interpolation or regularization framework. The proposed kernels appear to be particularly convenient for analyzing the e ffect of each (group of) variable(s) and computing sensitivity indices without recursivity.

Recommendations

• On ANOVA decompositions of kernels and Gaussian random field paths
• Metamodel construction for sensitivity analysis
• Global sensitivity analysis using support vector regression
• Analysis of variance on function spaces
• Functional ANOVA with multiple distributions: implications for the sensitivity analysis of computer experiments

Cited in

(18)

• Accelerating metabolic models evaluation with statistical metamodels: application to Salmonella infection models
• Bayesian quadrature, energy minimization, and space-filling design
• Sequential Learning of Active Subspaces
• Metamodel construction for sensitivity analysis
• Fast hyperbolic wavelet regression meets ANOVA
• Multifidelity information fusion algorithms for high-dimensional systems and massive data sets
• Polar Gaussian processes and experimental designs in circular domains
• Global sensitivity analysis with dependence measures
• On degeneracy and invariances of random fields paths with applications in Gaussian process modelling
• ANOVA decomposition of conditional Gaussian processes for sensitivity analysis with dependent inputs
• Sensitivity Prewarping for Local Surrogate Modeling
• Functional ANOVA with multiple distributions: implications for the sensitivity analysis of computer experiments
• On ANOVA decompositions of kernels and Gaussian random field paths
• Faster Kriging: facing high-dimensional simulators
• Efficient nonlocal linear image denoising: bilevel optimization with nonequispaced fast Fourier transform and matrix-free preconditioning
• High dimensional model representation constructed by support vector regression. I. Independent variables with known probability distributions
• Efficient learning of nonparametric directed acyclic graph with statistical guarantee
• Learning additive decompositions of multiattribute utility functions

This page was built for publication: ANOVA kernels and RKHS of zero mean functions for model-based sensitivity analysis

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1941425)