Mercer kernels and integrated variance experimental design: connections between Gaussian process regression and polynomial approximation
From MaRDI portal
Publication:5741201
DOI10.1137/15M1017119zbMATH Open1342.62061arXiv1503.00021OpenAlexW2964029376MaRDI QIDQ5741201FDOQ5741201
Authors: Alex Gorodetsky, Youssef M. Marzouk 
Publication date: 22 July 2016
Published in: SIAM/ASA Journal on Uncertainty Quantification (Search for Journal in Brave)
Abstract: This paper examines experimental design procedures used to develop surrogates of computational models, exploring the interplay between experimental designs and approximation algorithms. We focus on two widely used approximation approaches, Gaussian process (GP) regression and non-intrusive polynomial approximation. First, we introduce algorithms for minimizing a posterior integrated variance (IVAR) design criterion for GP regression. Our formulation treats design as a continuous optimization problem that can be solved with gradient-based methods on complex input domains, without resorting to greedy approximations. We show that minimizing IVAR in this way yields point sets with good interpolation properties, and that it enables more accurate GP regression than designs based on entropy minimization or mutual information maximization. Second, using a Mercer kernel/eigenfunction perspective on GP regression, we identify conditions under which GP regression coincides with pseudospectral polynomial approximation. Departures from these conditions can be understood as changes either to the kernel or to the experimental design itself. We then show how IVAR-optimal designs, while sacrificing discrete orthogonality of the kernel eigenfunctions, can yield lower approximation error than orthogonalizing point sets. Finally, we compare the performance of adaptive Gaussian process regression and adaptive pseudospectral approximation for several classes of target functions, identifying features that are favorable to the GP + IVAR approach.
Full work available at URL: https://arxiv.org/abs/1503.00021
Recommendations
- Cholesky-based experimental design for Gaussian process and kernel-based emulation and calibration
- Optimal designs for Gaussian process models via spectral decomposition
- Bayesian quadrature, energy minimization, and space-filling design
- scientific article; zbMATH DE number 7329270
- scientific article; zbMATH DE number 1209388
computer experimentsexperimental designGaussian process regressionuncertainty quantificationapproximation theorypolynomial approximationkernel interpolation
Nonparametric regression and quantile regression (62G08) Algorithms for approximation of functions (65D15)
Cites Work
- The design and analysis of computer experiments.
- Design and analysis of computer experiments. With comments and a rejoinder by the authors
- Gaussian processes for machine learning.
- Optimum experimental designs, with SAS
- Title not available (Why is that?)
- Real Analysis and Probability
- An Algorithm for the Construction of "D-Optimal" Experimental Designs
- Dimension-adaptive tensor-product quadrature
- On cardinal interpolation by Gaussian radial-basis functions: Properties of fundamental functions and estimates for Lebesgue constants
- Near-optimal data-independent point locations for radial basis function interpolation
- Positive integral operators in unbounded domains
- High dimensional polynomial interpolation on sparse grids
- Batch sequential designs for computer experiments
- Convergence of Unsymmetric Kernel‐Based Meshless Collocation Methods
- Title not available (Why is that?)
- Near-optimal sensor placements in Gaussian processes: theory, efficient algorithms and empirical studies
- Stability of kernel-based interpolation
- Title not available (Why is that?)
- Adaptive Smolyak Pseudospectral Approximations
- Sparse pseudospectral approximation method
- Uncertainty quantification given discontinuous model response and a limited number of model runs
- Asymptotic analysis of stochastic programs
- Characterization of discontinuities in high-dimensional stochastic problems on adaptive sparse grids
- Optimal designs for Gaussian process models via spectral decomposition
- Title not available (Why is that?)
- Spatial sampling design and covariance-robust minimax prediction based on convex design ideas
- Sequential Design with Mutual Information for Computer Experiments (MICE): Emulation of a Tsunami Model
- Mercer’s Theorem, Feature Maps, and Smoothing
- Title not available (Why is that?)
- Efficient Localization of Discontinuities in Complex Computational Simulations
- Title not available (Why is that?)
- An Exact Algorithm for Maximum Entropy Sampling
- Spectral approximation of the IMSE criterion for optimal designs in kernel-based interpolation models
- An extension of Mercer's theory to \(L^p\)
- Title not available (Why is that?)
Cited In (11)
- Gradient-based optimization for regression in the functional tensor-train format
- Locally induced Gaussian processes for large-scale simulation experiments
- Replication or Exploration? Sequential Design for Stochastic Simulation Experiments
- Adaptive Gaussian Process Approximation for Bayesian Inference with Expensive Likelihood Functions
- Optimal Off-line Experimentation for Games
- Solving Bayesian inverse problems with expensive likelihoods using constrained Gaussian processes and active learning
- Optimal experimental design: formulations and computations
- Clustered active-subspace based local Gaussian process emulator for high-dimensional and complex computer models
- Cholesky-Based Experimental Design for Gaussian Process and Kernel-Based Emulation and Calibration
- Bayesian Quadrature, Energy Minimization, and Space-Filling Design
- An Adaptive Minimum Spanning Tree Multielement Method for Uncertainty Quantification of Smooth and Discontinuous Responses
Uses Software
This page was built for publication: Mercer kernels and integrated variance experimental design: connections between Gaussian process regression and polynomial approximation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5741201)
