Gaussian process-based dimension reduction for goal-oriented sequential design

DOI10.1137/18M1167930MaRDI QIDQ4960979zbMATH OpenOpenAlexWikidataFDO

Authors Malek Ben Salem, François Bachoc, Lionel Tomaso, O. Roustant, Fabrice Gamboa

Publication date 24 April 2020

Published in SIAM/ASA Journal on Uncertainty Quantification (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1137/18m1167930

zbMATH Keywords

variable selectiondesign of experimentssurrogate modelingBayesian optimization

Mathematics Subject Classification ID

Bayesian inference (62F15) Gaussian processes (60G15) Statistical ranking and selection procedures (62F07) Optimal statistical designs (62K05) Sequential statistical design (62L05)

Recommendations

• An efficient dimension reduction for the Gaussian process emulation of two nested codes with functional outputs
• Sequential design of experiments for estimating percentiles of black-box functions
• Sequential design for achieving estimated accuracy of global sensitivities
• Symmetry results for decay solutions of elliptic systems in the whole space
• Sequential construction and dimension reduction of Gaussian processes under inequality constraints

Cites work

• A non-myopic utility function for statistical global optimization algorithms
• A taxonomy of global optimization methods based on response surfaces
• Asymptotic analysis of the role of spatial sampling for covariance parameter estimation of Gaussian processes
• Bayesian Design and Analysis of Computer Experiments: Use of Derivatives in Surface Prediction
• Bayesian optimization in a billion dimensions via random embeddings
• Cross validation and maximum likelihood estimations of hyper-parameters of Gaussian processes with model misspecification
• Cross-validation estimation of covariance parameters under fixed-domain asymptotics
• Derivative based global sensitivity measures and their link with global sensitivity indices
• Dimension Reduction for Gaussian Process Emulation: An Application to the Influence of Bathymetry on Tsunami Heights
• Efficient global optimization of expensive black-box functions
• Erratum: Active Subspace Methods in Theory and Practice: Applications to Kriging Surfaces
• Gaussian processes for machine learning.
• Global Sensitivity Analysis for Optimization with Variable Selection
• Gradient-based kernel dimension reduction for regression
• Hierarchical Nonlinear Approximation for Experimental Design and Statistical Data Fitting
• Interpolation of spatial data. Some theory for kriging
• Maximum Likelihood Estimation for a Smooth Gaussian Random Field Model
• New sensitivity analysis subordinated to a contrast
• Scattered Data Approximation
• Sequential design of computer experiments for the estimation of a probability of failure
• The design and analysis of computer experiments.
• Total interaction index: a variance-based sensitivity index for second-order interaction screening

Cited in

(2)

• Asymptotic properties of the maximum likelihood and cross validation estimators for transformed Gaussian processes
• The Zero Problem: Gaussian Process Emulators for Range-Constrained Computer Models

Describes a project that uses

Uses Software

• SMAC
• NLPQL
• DiceOptim
• DiceKriging
• EGO

This page was built for publication: Gaussian process-based dimension reduction for goal-oriented sequential design

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4960979)