Parameter estimation in high dimensional Gaussian distributions
From MaRDI portal
Publication:892469
DOI10.1007/S11222-012-9368-YzbMATH Open1325.62006arXiv1105.5256OpenAlexW2049766166MaRDI QIDQ892469FDOQ892469
D. Simpson, Jo Eidsvik, Erlend Aune
Publication date: 19 November 2015
Published in: Statistics and Computing (Search for Journal in Brave)
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.
Full work available at URL: https://arxiv.org/abs/1105.5256
Cites Work
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Title not available (Why is that?)
- Gaussian Markov Random Fields
- Algorithm 843
- mftoolbox
- Fixed Rank Kriging for Very Large Spatial Data Sets
- Functions of Matrices
- Approximate Bayesian inference for latent Gaussian models by using integrated nested Laplace approximations (with discussion)
- A Fast and High Quality Multilevel Scheme for Partitioning Irregular Graphs
- Fitting Gaussian Markov Random Fields to Gaussian Fields
- Approximate Likelihood for Large Irregularly Spaced Spatial Data
- Spatial models generated by nested stochastic partial differential equations, with an application to global ozone mapping
- An Explicit Link between Gaussian Fields and Gaussian Markov Random Fields: The Stochastic Partial Differential Equation Approach
- Foundations of time-frequency analysis
- Modifying a Sparse Cholesky Factorization
- Parallel Preconditioning with Sparse Approximate Inverses
- Some large-scale matrix computation problems
- Computing the Action of the Matrix Exponential, with an Application to Exponential Integrators
- A Stochastic Estimator of the Trace of the Influence Matrix for Laplacian Smoothing Splines
- Computing $A^\alpha, \log(A)$, and Related Matrix Functions by Contour Integrals
- A probing method for computing the diagonal of a matrix inverse
- Decay bounds and \(O(n)\) algorithms for approximating functions of sparse matrices
- An estimator for the diagonal of a matrix
- Fast Discrete Curvelet Transforms
- BiCGStab(\(\ell\)) for families of shifted linear systems
- Wavelet sparse approximate inverse preconditioners
- Bounds for the entries of matrix functions with applications to preconditioning
- A matrix-free approach for solving the parametric Gaussian process maximum likelihood problem
- Iterative numerical methods for sampling from high dimensional Gaussian distributions
Cited In (27)
- Multi-scale process modelling and distributed computation for spatial data
- Efficient Covariance Approximations for Large Sparse Precision Matrices
- Title not available (Why is that?)
- Relaxing the Gaussian assumption in shrinkage and SURE in high dimension
- Locally induced Gaussian processes for large-scale simulation experiments
- Linear-Cost Covariance Functions for Gaussian Random Fields
- A general scheme for log-determinant computation of matrices via stochastic polynomial approximation
- Comments on: ``Comparing and selecting spatial predictors using local criteria
- Decay Bounds for Functions of Hermitian Matrices with Banded or Kronecker Structure
- Localization in Matrix Computations: Theory and Applications
- Estimating Orthant Probabilities of High-Dimensional Gaussian Vectors with An Application to Set Estimation
- Fast Estimation of $tr(f(A))$ via Stochastic Lanczos Quadrature
- Fast Spatial Gaussian Process Maximum Likelihood Estimation via Skeletonization Factorizations
- Fitting large-scale structured additive regression models using Krylov subspace methods
- Randomized block Krylov subspace methods for trace and log-determinant estimators
- Regularized parameter estimation of high dimensional distribution
- Computing tail areas for a high-dimensional Gaussian mixture
- Large-scale local surrogate modeling of stochastic simulation experiments
- Iterative numerical methods for sampling from high dimensional Gaussian distributions
- Analysis of stochastic probing methods for estimating the trace of functions of sparse symmetric matrices
- Spatial regression with non-parametric modeling of Fourier coefficients
- Stochastic approximation of score functions for Gaussian processes
- Optimal estimation of high-dimensional Gaussian location mixtures
- Approximating Spectral Sums of Large-Scale Matrices using Stochastic Chebyshev Approximations
- A determinant-free method to simulate the parameters of large Gaussian fields
- Kryging: geostatistical analysis of large-scale datasets using Krylov subspace methods
- On Russian roulette estimates for Bayesian inference with doubly-intractable likelihoods
Uses Software
Recommendations
- Title not available (Why is that?) 👍 👎
- Parameter Estimation For Multivariate Generalized Gaussian Distributions 👍 👎
- Gaussian approximations in high dimensional estimation 👍 👎
- Estimation in High-Dimensional Analysis and Multivariate Linear Models 👍 👎
- Regularized parameter estimation of high dimensional distribution 👍 👎
- Bayesian estimation in a high dimensional parameter framework 👍 👎
- Estimation of parameters of homogeneous Gaussian random fields 👍 👎
- Regularized Parameter Estimation in High-Dimensional Gaussian Mixture Models 👍 👎
- Estimation of Parameters in Gaussian Hypergeometric Distributions 👍 👎
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)
