A non-intrusive solution to the ill-conditioning problem of the gradient-enhanced Gaussian covariance matrix for Gaussian processes

DOI10.1007/s10915-023-02190-wOpenAlexW4366122358MaRDI QIDQ6101651

André L. Marchildon, David W. Zingg

Publication date: 20 June 2023

Published in: Journal of Scientific Computing (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1007/s10915-023-02190-w

zbMATH Keywords

Gaussian processcondition numbercovariance matrixBayesian optimization

Mathematics Subject Classification ID

Gaussian processes (60G15) Probabilistic models, generic numerical methods in probability and statistics (65C20) Numerical optimization and variational techniques (65K10)

Cites Work

• Unnamed Item
• Unnamed Item
• On the condition number anomaly of Gaussian correlation matrices
• Cases for the nugget in modeling computer experiments
• A tutorial on Gaussian process regression: modelling, exploring, and exploiting functions
• The effect of the nugget on Gaussian process emulators of computer models
• Exploiting Hessian matrix and trust-region algorithm in hyperparameters estimation of Gaussian process
• On the Instability Issue of Gradient-Enhanced Gaussian Process Emulators for Computer Experiments
• Solving Systems of Linear Equations With a Positive Definite, Symmetric, but Possibly Ill-Conditioned Matrix
• An adjoint for likelihood maximization
• On the Maximum Likelihood Training of Gradient-Enhanced Spatial Gaussian Processes
• A comparative evaluation of genetic and gradient-based algorithms applied to aerodynamic optimization
• On the condition number of covariance matrices in kriging, estimation, and simulation of random fields