Physics-informed cokriging: a Gaussian-process-regression-based multifidelity method for data-model convergence
From MaRDI portal
Publication:2222351
Abstract: In this work, we propose a new Gaussian process regression (GPR)-based multifidelity method: physics-informed CoKriging (CoPhIK). In CoKriging-based multifidelity methods, the quantities of interest are modeled as linear combinations of multiple parameterized stationary Gaussian processes (GPs), and the hyperparameters of these GPs are estimated from data via optimization. In CoPhIK, we construct a GP representing low-fidelity data using physics-informed Kriging (PhIK), and model the discrepancy between low- and high-fidelity data using a parameterized GP with hyperparameters identified via optimization. Our approach reduces the cost of optimization for inferring hyperparameters by incorporating partial physical knowledge. We prove that the physical constraints in the form of deterministic linear operators are satisfied up to an error bound. Furthermore, we combine CoPhIK with a greedy active learning algorithm for guiding the selection of additional observation locations. The efficiency and accuracy of CoPhIK are demonstrated for reconstructing the partially observed modified Branin function, reconstructing the sparsely observed state of a steady state heat transport problem, and learning a conservative tracer distribution from sparse tracer concentration measurements.
Recommendations
- Physics information aided kriging using stochastic simulation models
- Multifidelity Data Fusion via Gradient-Enhanced Gaussian Process Regression
- Extended co-Kriging interpolation method based on multi-fidelity data
- Theoretical investigations of the new cokriging method for variable-fidelity surrogate modeling. Well-posedness and maximum likelihood training.
- APIK: Active Physics-Informed Kriging Model with Partial Differential Equations
Cites work
- A class of bivariate heavy-tailed distributions
- Adaptive ANOVA decomposition of stochastic incompressible and compressible flows
- Asymptotic analysis of the learning curve for Gaussian process regression
- Bayesian solution uncertainty quantification for differential equations
- Design and analysis of computer experiments. With comments and a rejoinder by the authors
- Efficient global optimization of expensive black-box functions
- Gaussian process for nonstationary time series prediction
- Gaussian processes for machine learning.
- High-Order Collocation Methods for Differential Equations with Random Inputs
- scientific article; zbMATH DE number 53679 (Why is no real title available?)
- scientific article; zbMATH DE number 1304257 (Why is no real title available?)
- scientific article; zbMATH DE number 1149423 (Why is no real title available?)
- Machine learning of linear differential equations using Gaussian processes
- Multi-fidelity optimization via surrogate modelling
- Multi-fidelity stochastic collocation method for computation of statistical moments
- Multilevel Monte Carlo Path Simulation
- Near-optimal sensor placements in Gaussian processes: theory, efficient algorithms and empirical studies
- Nonlinear information fusion algorithms for data-efficient multi-fidelity modelling
- Numerical Gaussian processes for time-dependent and nonlinear partial differential equations
- Predicting the output from a complex computer code when fast approximations are available
- Probabilistic numerics and uncertainty in computations
- Probability and cumulative density function methods for the stochastic advection-reaction equation
- Reweighted \(\ell_1\) minimization method for stochastic elliptic differential equations
- Support vector machine active learning with applications to text classification
Cited in
(21)- Theoretical investigations of the new cokriging method for variable-fidelity surrogate modeling. Well-posedness and maximum likelihood training.
- Machine learning for metal additive manufacturing: predicting temperature and melt pool fluid dynamics using physics-informed neural networks
- Monte Carlo fPINNs: deep learning method for forward and inverse problems involving high dimensional fractional partial differential equations
- Conditional Karhunen-Loève expansion for uncertainty quantification and active learning in partial differential equation models
- Physics-informed machine learning with conditional Karhunen-Loève expansions
- Deep coregionalization for the emulation of simulation-based spatial-temporal fields
- Augmented Gaussian random field: theory and computation
- Active learning with multifidelity modeling for efficient rare event simulation
- Physics information aided kriging using stochastic simulation models
- APIK: Active Physics-Informed Kriging Model with Partial Differential Equations
- Multifidelity Data Fusion via Gradient-Enhanced Gaussian Process Regression
- When Bifidelity Meets CoKriging: An Efficient Physics-Informed MultiFidelity Method
- Physics-informed generative adversarial networks for stochastic differential equations
- Subset simulation for probabilistic computer models
- Improved training of physics-informed neural networks for parabolic differential equations with sharply perturbed initial conditions
- Stochastic PDE representation of random fields for large-scale Gaussian process regression and statistical finite element analysis
- A hybrid data-driven-physics-constrained Gaussian process regression framework with deep kernel for uncertainty quantification
- Physics-informed machine learning method with space-time Karhunen-Loève expansions for forward and inverse partial differential equations
- Physics-informed genetic programming for discovery of partial differential equations from scarce and noisy data
- Physics-informed solution reconstruction in elasticity and heat transfer using the explicit constraint force method
- Physics-based linear regression for high-dimensional forward uncertainty quantification
This page was built for publication: Physics-informed cokriging: a Gaussian-process-regression-based multifidelity method for data-model convergence
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2222351)
