Augmented Gaussian random field: theory and computation

DOI10.3934/DCDSS.2021098MaRDI QIDQ2129158zbMATH OpenOpenAlexFDO

Authors Sheng Zhang, Xiu Yang, S. Tindel, Guang Lin

Publication date 22 April 2022

Published in Discrete and Continuous Dynamical Systems. Series S (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2009.01961

zbMATH Keywords

Gaussian random fieldmissing dataGaussian process regressionnoisy dataarbitrary order derivatives

Mathematics Subject Classification ID

Gaussian processes (60G15) Random fields (60G60)

Abstract: We propose the novel augmented Gaussian random field (AGRF), which is a universal framework incorporating the data of observable and derivatives of any order. Rigorous theory is established. We prove that under certain conditions, the observable and its derivatives of any order are governed by a single Gaussian random field, which is the aforementioned AGRF. As a corollary, the statement ``the derivative of a Gaussian process remains a Gaussian process is validated, since the derivative is represented by a part of the AGRF. Moreover, a computational method corresponding to the universal AGRF framework is constructed. Both noiseless and noisy scenarios are considered. Formulas of the posterior distributions are deduced in a nice closed form. A significant advantage of our computational method is that the universal AGRF framework provides a natural way to incorporate arbitrary order derivatives and deal with missing data. We use four numerical examples to demonstrate the effectiveness of the computational method. The numerical examples are composite function, damped harmonic oscillator, Korteweg-De Vries equation, and Burgers' equation.

Recommendations

• scientific article; zbMATH DE number 4155584
• A note on generalized Gaussian random fields
• Gaussian processes and Gaussian random fields
• scientific article; zbMATH DE number 3846589
• Advances in Gaussian random field generation: a review
• Gaussian Markov Random Fields
• scientific article; zbMATH DE number 194255
• Optimization of Gaussian random fields
• Gaussian approximation of mixing random fields
• scientific article; zbMATH DE number 1795881

Cites work

• Bayesian Design and Analysis of Computer Experiments: Use of Derivatives in Surface Prediction
• Bayesian Probabilistic Numerical Methods
• Bayesian monotone regression using Gaussian process projection
• Bayesian solution uncertainty quantification for differential equations
• Design and analysis of computer experiments. With comments and a rejoinder by the authors
• Finite-Dimensional Gaussian Approximation with Linear Inequality Constraints
• Gaussian process modeling with inequality constraints
• Gaussian processes for machine learning.
• Machine learning of linear differential equations using Gaussian processes
• Multi-fidelity optimization via surrogate modelling
• Multifidelity Data Fusion via Gradient-Enhanced Gaussian Process Regression
• Nonlinear information fusion algorithms for data-efficient multi-fidelity modelling
• Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
• Physics-informed cokriging: a Gaussian-process-regression-based multifidelity method for data-model convergence
• Predicting the output from a complex computer code when fast approximations are available
• Probabilistic numerics and uncertainty in computations
• RECURSIVE CO-KRIGING MODEL FOR DESIGN OF COMPUTER EXPERIMENTS WITH MULTIPLE LEVELS OF FIDELITY
• Real Analysis and Probability
• The Geometry of Random Fields
• Theory of probability and random processes.
• When Bifidelity Meets CoKriging: An Efficient Physics-Informed MultiFidelity Method

Cited in

(2)

• Augmentation schemes for particle MCMC
• A kernel framework for learning differential equations and their solution operators

This page was built for publication: Augmented Gaussian random field: theory and computation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2129158)