Parameter estimation in high dimensional Gaussian distributions
From MaRDI portal
Publication:892469
Abstract: In order to compute the log-likelihood for high dimensional spatial Gaussian models, it is necessary to compute the determinant of the large, sparse, symmetric positive definite precision matrix, Q. Traditional methods for evaluating the log-likelihood for very large models may fail due to the massive memory requirements. We present a novel approach for evaluating such likelihoods when the matrix-vector product, Qv, is fast to compute. In this approach we utilise matrix functions, Krylov subspaces, and probing vectors to construct an iterative method for computing the log-likelihood.
Recommendations
- Gaussian approximations in high dimensional estimation
- Regularized parameter estimation of high dimensional distribution
- Estimation of Parameters in Gaussian Hypergeometric Distributions
- Bayesian estimation in a high dimensional parameter framework
- Parameter Estimation For Multivariate Generalized Gaussian Distributions
- Regularized parameter estimation in high-dimensional Gaussian mixture models
- scientific article; zbMATH DE number 192974
- Estimation in high-dimensional analysis and multivariate linear models
- Estimation of parameters of homogeneous Gaussian random fields
Cites work
- scientific article; zbMATH DE number 3513115 (Why is no real title available?)
- scientific article; zbMATH DE number 1083101 (Why is no real title available?)
- scientific article; zbMATH DE number 1953444 (Why is no real title available?)
- scientific article; zbMATH DE number 1380579 (Why is no real title available?)
- scientific article; zbMATH DE number 806799 (Why is no real title available?)
- scientific article; zbMATH DE number 2222870 (Why is no real title available?)
- A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs
- A Stochastic Estimator of the Trace of the Influence Matrix for Laplacian Smoothing Splines
- A matrix-free approach for solving the parametric Gaussian process maximum likelihood problem
- A probing method for computing the diagonal of a matrix inverse.
- Algorithm 843
- An estimator for the diagonal of a matrix
- An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach
- Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations (with discussion)
- Approximate Likelihood for Large Irregularly Spaced Spatial Data
- BiCGStab(\(\ell\)) for families of shifted linear systems
- Bounds for the entries of matrix functions with applications to preconditioning
- Computing $A^\alpha, \log(A)$, and Related Matrix Functions by Contour Integrals
- Computing the action of the matrix exponential, with an application to exponential integrators
- Decay bounds and \(O(n)\) algorithms for approximating functions of sparse matrices
- Fast Discrete Curvelet Transforms
- Fitting Gaussian Markov Random Fields to Gaussian Fields
- Fixed Rank Kriging for Very Large Spatial Data Sets
- Foundations of time-frequency analysis
- Functions of Matrices
- Gaussian Markov Random Fields
- Iterative numerical methods for sampling from high dimensional Gaussian distributions
- Modifying a Sparse Cholesky Factorization
- Parallel Preconditioning with Sparse Approximate Inverses
- Some large-scale matrix computation problems
- Spatial models generated by nested stochastic partial differential equations, with an application to global ozone mapping
- Wavelet sparse approximate inverse preconditioners
- mftoolbox
Cited in
(28)- Comments on: ``Comparing and selecting spatial predictors using local criteria
- Analysis of stochastic probing methods for estimating the trace of functions of sparse symmetric matrices
- Approximating spectral sums of large-scale matrices using stochastic Chebyshev approximations
- Large-scale local surrogate modeling of stochastic simulation experiments
- Spatial regression with non-parametric modeling of Fourier coefficients
- Fitting large-scale structured additive regression models using Krylov subspace methods
- Randomized block Krylov subspace methods for trace and log-determinant estimators
- Multi-scale process modelling and distributed computation for spatial data
- Linear-Cost Covariance Functions for Gaussian Random Fields
- Iterative numerical methods for sampling from high dimensional Gaussian distributions
- Fast estimation of \(\mathrm{tr}(f(A))\) via stochastic Lanczos quadrature
- Decay bounds for functions of Hermitian matrices with banded or Kronecker structure
- Stochastic approximation of score functions for Gaussian processes
- Estimating Orthant Probabilities of High-Dimensional Gaussian Vectors with An Application to Set Estimation
- Optimal estimation of high-dimensional Gaussian location mixtures
- Localization in matrix computations: theory and applications
- scientific article; zbMATH DE number 7049740 (Why is no real title available?)
- Efficient Covariance Approximations for Large Sparse Precision Matrices
- A determinant-free method to simulate the parameters of large Gaussian fields
- Kryging: geostatistical analysis of large-scale datasets using Krylov subspace methods
- Regularized parameter estimation of high dimensional distribution
- Fast spatial Gaussian process maximum likelihood estimation via skeletonization factorizations
- On Russian roulette estimates for Bayesian inference with doubly-intractable likelihoods
- Relaxing the Gaussian assumption in shrinkage and SURE in high dimension
- Computing tail areas for a high-dimensional Gaussian mixture
- Log-det approximation based on uniformly distributed seeds and its application to Gaussian process regression
- Locally induced Gaussian processes for large-scale simulation experiments
- A general scheme for log-determinant computation of matrices via stochastic polynomial approximation
Describes a project that uses
Uses Software
This page was built for publication: Parameter estimation in high dimensional Gaussian distributions
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q892469)
