A comparison between Markov approximations and other methods for large spatial data sets

DOI10.1016/J.CSDA.2012.11.011zbMATH Open1349.62445DBLPjournals/csda/BolinL13arXiv1106.1980OpenAlexW2158359474WikidataQ57266351 ScholiaQ57266351MaRDI QIDQ333683FDOQ333683

Finn Lindgren, David Bolin

Publication date: 31 October 2016

Published in: Computational Statistics and Data Analysis (Search for Journal in Brave)

Abstract: The Mat'ern covariance function is a popular choice for modeling dependence in spatial environmental data. Standard Mat'ern covariance models are, however, often computationally infeasible for large data sets. In this work, recent results for Markov approximations of Gaussian Mat'{e}rn fields based on Hilbert space approximations are extended using wavelet basis functions. These Markov approximations are compared with two of the most popular methods for efficient covariance approximations; covariance tapering and the process convolution method. The results show that, for a given computational cost, the Markov methods have a substantial gain in accuracy compared with the other methods.

Full work available at URL: https://arxiv.org/abs/1106.1980

zbMATH Keywords

waveletskriging[https://portal.mardi4nfdi.de/w/index.php?title=+Special%3ASearch&search=Mat%EF%BF%BD%EF%BF%BDrn+covariances&go=Go Mat��rn covariances]Markov random fieldscovariance taperingprocess convolutions

Mathematics Subject Classification ID

Inference from spatial processes (62M30) Random fields; image analysis (62M40)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Gaussian Markov Random Fields
• Fixed Rank Kriging for Very Large Spatial Data Sets
• Interpolation of spatial data. Some theory for kriging
• Compactly supported correlation functions
• Piecewise polynomial, positive definite and compactly supported radial functions of minimal degree
• Statistical Methods for Spatial Data Analysis
• Gaussian Predictive Process Models for Large Spatial Data Sets
• Fitting Gaussian Markov Random Fields to Gaussian Fields
• Spatial models generated by nested stochastic partial differential equations, with an application to global ozone mapping
• Inconsistent Estimation and Asymptotically Equal Interpolations in Model-Based Geostatistics
• An Explicit Link between Gaussian Fields and Gaussian Markov Random Fields: The Stochastic Partial Differential Equation Approach
• A comparative study of Gaussian geostatistical models and Gaussian Markov random field models
• Handbook of Spatial Statistics
• Ten Lectures on Wavelets
• Multiresolution models for nonstationary spatial covariance functions
• On Compactly Supported Spline Wavelets and a Duality Principle
• Approximate Bayesian inference for large spatial datasets using predictive process models
• A Class of Convolution-Based Models for Spatio-Temporal Processes with Non-Separable Covariance Structure
• Calibrated spatial moving average simulations

Cited In (10)

• On the choice of the mesh for the analysis of geostatistical data using R-INLA
• Beyond the valley of the covariance function
• Spatial Matérn fields driven by non-Gaussian noise
• Distributed Bayesian inference in massive spatial data
• Regularity theory for a new class of fractional parabolic stochastic evolution equations
• Process convolution approaches for modeling interacting trajectories
• An ensemble solver for segregated cardiovascular FSI
• A Bayesian General Linear Modeling Approach to Cortical Surface fMRI Data Analysis
• Sparse Approximations of Fractional Matérn Fields
• Finite Element Representations of Gaussian Processes: Balancing Numerical and Statistical Accuracy

Uses Software

• GMRFLib
• FRK

This page was built for publication: A comparison between Markov approximations and other methods for large spatial data sets

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q333683)