A framework for controlling sources of inaccuracy in Gaussian process emulation of deterministic computer experiments

DOI10.1137/17M1131210zbMATH Open1403.62143arXiv1411.7049OpenAlexW2963732218MaRDI QIDQ3176230FDOQ3176230

Authors: Benjamin Haaland, Wenjia Wang, Vaibhav Maheshwari 

Publication date: 19 July 2018

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

Abstract: Computer experiments have become ubiquitous in science and engineering. Commonly, runs of these simulations demand considerable time and computing, making experimental design extremely important in gaining high quality information with limited time and resources. Principles of experimental design are proposed and justified which ensure high nominal, numeric, and parameter estimation accuracy for Gaussian process emulation of deterministic simulations. The space-filling properties "small fill distance" and "large separation distance" are only weakly conflicting and ensure well-controlled nominal, numeric, and parameter estimation error, while non-stationarity requires a greater density of experimental inputs in regions of the input space with more quickly decaying correlation. This work will provide scientists and engineers with robust, rigorously justified, and practically useful overarching principles for selecting combinations of simulation inputs with high information content.

Full work available at URL: https://arxiv.org/abs/1411.7049

Recommendations

• Computer experiment designs for accurate prediction
• Design and analysis of computer experiments. With comments and a rejoinder by the authors
• scientific article; zbMATH DE number 1209388
• Analysis methods for computer experiments: how to assess and what counts?
• Computer emulation with nonstationary Gaussian processes

zbMATH Keywords

experimental designinterpolationreproducing kernel Hilbert spaceGaussian processcomputer experimentemulation

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Gaussian processes (60G15) Optimal statistical designs (62K05) Design of statistical experiments (62K99)

Cites Work

• Efficient global optimization of expensive black-box functions
• Numerical Optimization
• The design and analysis of computer experiments.
• Design and analysis of computer experiments. With comments and a rejoinder by the authors
• Theory of Reproducing Kernels
• A Comparison of Three Methods for Selecting Values of Input Variables in the Analysis of Output from a Computer Code
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Maximum projection designs for computer experiments
• Composite Gaussian process models for emulating expensive functions
• Accurate emulators for large-scale computer experiments
• Title not available (Why is that?)
• Scattered Data Approximation
• Orthogonal Array-Based Latin Hypercubes
• Title not available (Why is that?)
• Title not available (Why is that?)
• Spatial sampling design for parameter estimation of the covariance function
• THE DESIGN OF OPTIMUM MULTIFACTORIAL EXPERIMENTS
• Title not available (Why is that?)
• Experimental design in a class of models
• Title not available (Why is that?)
• Error estimates and condition numbers for radial basis function interpolation
• Homotopies for computation of fixed points
• Orthogonal and nearly orthogonal designs for computer experiments
• An efficient algorithm for constructing optimal design of computer experiments
• Design of computer experiments: space filling and beyond
• Average case complexity of linear multivariate problems. I: Theory
• Maximin design on non hypercube domains and kernel interpolation
• On the Efficiency of Experimental Designs
• Reverses of Schwarz inequality in inner product spaces with applications
• Title not available (Why is that?)

Cited In (8)

• Title not available (Why is that?)
• Error estimation properties of Gaussian process models in stochastic simulations
• Flexible Correlation Structure for Accurate Prediction and Uncertainty Quantification in Bayesian Gaussian Process Emulation of a Computer Model
• Controlling Sources of Inaccuracy in Stochastic Kriging
• The effect of the nugget on Gaussian process emulators of computer models
• Fast Calibration for Computer Models with Massive Physical Observations
• Efficient kriging using interleaved lattice-based designs with low fill and high separation distance properties
• On the interface between nested designs and the multi-step interpolator

Uses Software

• EGO
• mlegp

This page was built for publication: A framework for controlling sources of inaccuracy in Gaussian process emulation of deterministic computer experiments

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3176230)