Finite-Dimensional Gaussian Approximation with Linear Inequality Constraints
From MaRDI portal
Abstract: Introducing inequality constraints in Gaussian process (GP) models can lead to more realistic uncertainties in learning a great variety of real-world problems. We consider the finite-dimensional Gaussian approach from Maatouk and Bay (2017) which can satisfy inequality conditions everywhere (either boundedness, monotonicity or convexity). Our contributions are threefold. First, we extend their approach in order to deal with general sets of linear inequalities. Second, we explore several Markov Chain Monte Carlo (MCMC) techniques to approximate the posterior distribution. Third, we investigate theoretical and numerical properties of the constrained likelihood for covariance parameter estimation. According to experiments on both artificial and real data, our full framework together with a Hamiltonian Monte Carlo-based sampler provides efficient results on both data fitting and uncertainty quantification.
Recommendations
- Finite-dimensional approximation of Gaussian processes with linear inequality constraints and noisy observations
- Gaussian processes with linear operator inequality constraints
- scientific article; zbMATH DE number 1304258
- Approximation, dimension reduction, and nonconvex optimization using linear superpositions of Gaussians
- scientific article; zbMATH DE number 854127
- A technique for obtaining approximate solutions in Gaussian equivalent linearization
- scientific article; zbMATH DE number 79269
- scientific article; zbMATH DE number 3890597
- scientific article; zbMATH DE number 175883
- A gradient formula for linear chance constraints under Gaussian distribution
Cites work
- A class of globally convergent optimization methods based on conservative convex separable approximations
- A New Rejection Sampling Method for Truncated Multivariate Gaussian Random Variables Restricted to Convex Sets
- A supermartingale approach to Gaussian process based sequential design of experiments
- Approximation, metric entropy and small ball estimates for Gaussian measures
- Approximations for binary Gaussian process classification
- Bayesian learning for neural networks
- Estimating structured correlation matrices in smooth Gaussian random field models.
- Fixed-domain asymptotic properties of tapered maximum likelihood estimators
- Fixed-domain asymptotics for a subclass of Matérn-type Gaussian random fields
- Gaussian process emulators for computer experiments with inequality constraints
- Gaussian process modeling with inequality constraints
- Generalization of the Kimeldorf-Wahba correspondence for constrained interpolation
- Handbook of Markov Chain Monte Carlo
- scientific article; zbMATH DE number 1713116 (Why is no real title available?)
- scientific article; zbMATH DE number 3858094 (Why is no real title available?)
- scientific article; zbMATH DE number 51414 (Why is no real title available?)
- scientific article; zbMATH DE number 3608897 (Why is no real title available?)
- scientific article; zbMATH DE number 1460605 (Why is no real title available?)
- Inconsistent Estimation and Asymptotically Equal Interpolations in Model-Based Geostatistics
- Interpolation of spatial data. Some theory for kriging
- Kriging of financial term-structures
- Level Sets and Extrema of Random Processes and Fields
- Machine learning. A probabilistic perspective
- Maximum likelihood estimation of parameters under a spatial sampling scheme
- Modelling and control of dynamic systems using Gaussian process models
- Monotone Emulation of Computer Experiments
- Multivariate output analysis for Markov chain Monte Carlo
- Sampling constrained probability distributions using spherical augmentation
- Simulating normalizing constants: From importance sampling to bridge sampling to path sampling
- The design and analysis of computer experiments.
- The normal law under linear restrictions: simulation and estimation via minimax tilting
Cited in
(30)- Gaussian process modeling with inequality constraints
- Composite likelihood estimation for a Gaussian process under fixed domain asymptotics
- Gaussian process optimization with failures: classification and convergence proof
- A role for symmetry in the Bayesian solution of differential equations
- Augmented Gaussian random field: theory and computation
- Adaptive design for Gaussian process regression under censoring
- Fixed-domain asymptotic properties of maximum composite likelihood estimators for Gaussian processes
- Some statistical issues in climate science
- Maximum likelihood estimation for Gaussian processes under inequality constraints
- scientific article; zbMATH DE number 5356341 (Why is no real title available?)
- Sequential construction and dimension reduction of Gaussian processes under inequality constraints
- Approximating Gaussian Process Emulators with Linear Inequality Constraints and Noisy Observations via MC and MCMC
- Gaussian processes with linear operator inequality constraints
- Asymptotic Analysis of Maximum Likelihood Estimation of Covariance Parameters for Gaussian Processes: An Introduction with Proofs
- Finite-dimensional approximation of Gaussian processes with linear inequality constraints and noisy observations
- Convex optimization of nonlinear inequality with higher order derivatives
- The Zero Problem: Gaussian Process Emulators for Range-Constrained Computer Models
- Knowledge-based modeling of simulation behavior for Bayesian optimization
- Nonlinear response modelling of material systems using constrained Gaussian processes
- Block-additive Gaussian processes under monotonicity constraints
- Bayesian analysis of constrained Gaussian processes
- Estimation of the means for k multivariate normal populations based on simple partial order restriction
- Error bounds for a kernel-based constrained optimal smoothing approximation
- Equivalence between constrained optimal smoothing and Bayesian estimation
- Large-scale constrained Gaussian processes for shape-restricted function estimation
- Posterior contraction rates for constrained deep Gaussian processes in density estimation and classification
- Efficient monotonic Gaussian processes via randomize-then-optimize
- Gaussian process with dissolution spline kernel for in vitro dissolution testing
- Relaxed Gaussian process interpolation: a goal-oriented approach to Bayesian optimization
- Adaptive grid designs for classifying monotonic binary deterministic computer simulations
This page was built for publication: Finite-Dimensional Gaussian Approximation with Linear Inequality Constraints
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4611516)
