%0 Journal Article %1 eidsvik_approximate_2012 %A Eidsvik, Jo %A Finley, Andrew O. %A Banerjee, Sudipto %A Rue, Håvard %D 2012 %J Computational Statistics & Data Analysis %K Approximate Bayesian Laplace Spatial \_tablet\_modified, approximation, based geostatistic, inference, modelling, process processes %N 6 %P 1362--1380 %R 10.1016/j.csda.2011.10.022 %T Approximate Bayesian inference for large spatial datasets using predictive process models %U http://www.sciencedirect.com/science/article/pii/S0167947311003938 %V 56 %X The challenges of estimating hierarchical spatial models to large datasets are addressed. With the increasing availability of geocoded scientific data, hierarchical models involving spatial processes have become a popular method for carrying out spatial inference. Such models are customarily estimated using Markov chain Monte Carlo algorithms that, while immensely flexible, can become prohibitively expensive. In particular, fitting hierarchical spatial models often involves expensive decompositions of dense matrices whose computational complexity increases in cubic order with the number of spatial locations. Such matrix computations are required in each iteration of the Markov chain Monte Carlo algorithm, rendering them infeasible for large spatial datasets. The computational challenges in analyzing large spatial datasets are considered by merging two recent developments. First, the predictive process model is used as a reduced-rank spatial process, to diminish the dimensionality of the model. Then a computational framework is developed for estimating predictive process models using the integrated nested Laplace approximation. The settings where the first stage likelihood is Gaussian or non-Gaussian are discussed. Issues such as predictions and model comparisons are also discussed. Results are presented for synthetic data and several environmental datasets.

@article{eidsvik_approximate_2012, abstract = {The challenges of estimating hierarchical spatial models to large datasets are addressed. With the increasing availability of geocoded scientific data, hierarchical models involving spatial processes have become a popular method for carrying out spatial inference. Such models are customarily estimated using Markov chain Monte Carlo algorithms that, while immensely flexible, can become prohibitively expensive. In particular, fitting hierarchical spatial models often involves expensive decompositions of dense matrices whose computational complexity increases in cubic order with the number of spatial locations. Such matrix computations are required in each iteration of the Markov chain Monte Carlo algorithm, rendering them infeasible for large spatial datasets. The computational challenges in analyzing large spatial datasets are considered by merging two recent developments. First, the predictive process model is used as a reduced-rank spatial process, to diminish the dimensionality of the model. Then a computational framework is developed for estimating predictive process models using the integrated nested Laplace approximation. The settings where the first stage likelihood is Gaussian or non-Gaussian are discussed. Issues such as predictions and model comparisons are also discussed. Results are presented for synthetic data and several environmental datasets.}, added-at = {2017-01-09T13:57:26.000+0100}, author = {Eidsvik, Jo and Finley, Andrew O. and Banerjee, Sudipto and Rue, Håvard}, biburl = {https://www.bibsonomy.org/bibtex/2c2b17a43fb49ea20241ac9a1a07b760b/yourwelcome}, doi = {10.1016/j.csda.2011.10.022}, interhash = {3f4030f8b01953dbb37a023fe53213d1}, intrahash = {c2b17a43fb49ea20241ac9a1a07b760b}, issn = {0167-9473}, journal = {Computational Statistics \& Data Analysis}, keywords = {Approximate Bayesian Laplace Spatial \_tablet\_modified, approximation, based geostatistic, inference, modelling, process processes}, month = jun, number = 6, pages = {1362--1380}, timestamp = {2017-01-09T14:01:11.000+0100}, title = {Approximate {Bayesian} inference for large spatial datasets using predictive process models}, url = {http://www.sciencedirect.com/science/article/pii/S0167947311003938}, urldate = {2014-02-03}, volume = 56, year = 2012 }