Hierarchical Models for Spatial Data with Errors that are Correlated with the Latent Process

  title={Hierarchical Models for Spatial Data with Errors that are Correlated with the Latent Process},
  author={Jonathan R. Bradley and Christopher K. Wikle and Scott H. Holan},
  journal={Statistica Sinica},
Prediction of a spatial process using a “big dataset” has become a topical area of research over the last decade. The available solutions often involve placing strong assumptions on the error process associated with the data. Specifically, it has typically been assumed that the data are equal to the spatial process of principal interest plus a mutually independent error process. This is done to avoid modeling confounded cross-covariances between the signal and noise within an additive model. In… 

Figures from this paper

Bayesian inference for big spatial data using non-stationary spectral simulation

A Fused Gaussian Process Model for Very Large Spatial Data

The results demonstrate the computational and inferential benefits of FGP over competing methods and show that FGP is robust against model misspecification and captures spatial nonstationarity.

Modeling Dependence in Spatio-Temporal Econometrics

This chapter illustrates how noisy and missing data can be accounted for by using the STAR-like models as process models, alongside a data model and potentially a parameter model, in a hierarchical statistical model (HM).

An Approach to Incorporate Subsampling Into a Generic Bayesian Hierarchical Model

  • J. Bradley
  • Computer Science
    J. Comput. Graph. Stat.
  • 2021
A "data subset model" is introduced to the popular “data model, process model, and parameter model” framework used to summarize Bayesian hierarchical models, and several properties of the data subset model are provided including: propriety, partial sufficiency, and semi-parametric properties.

Joint Bayesian Analysis of Multiple Response-Types Using the Hierarchical Generalized Transformation Model

A reasonable easy-to-implement all-purpose method that “converts” a Bayesian statistical model for continuous responses into aBayesian model for multiple response-type datasets and demonstrates the flexibility of the model through an application to Bayesian additive regression trees (BART) and a spatio-temporal mixed effects model.

Latent Multivariate Log-Gamma Models for High-Dimensional Multi-Type Responses with Application to Daily Fine Particulate Matter and Mortality Counts

A Bayesian hierarchical model is proposed for high-dimensional "multi-type" responses that capitalizes on the shared conjugate structure between the Weibull, Poisson, and multivariate log-gamma distributions, and uses dimension reduction to aid with computation.

Joint spatio-temporal analysis of multiple response types using the hierarchical generalized transformation model with application to coronavirus disease 2019 and social distancing

A reasonable easy-to-implement all-purpose method that "converts" a statistical model for continuous responses (the preferred model) into a Bayesian model for multi-response data sets for social distancing purposes.

Effects of a Government-Academic Partnership: Has the NSF-CENSUS Bureau Research Network Helped Improve the US Statistical System?

The National Science Foundation-Census Bureau Research Network (NCRN) was established in 2011 to create interdisciplinary research nodes on methodological questions of interest and significance to

Bayesian analysis of count-valued, binary-valued, and continuous-valued responses using unknown transformations

A reasonable easy-to-implement all-purpose method that "converts" a Bayesian statistical model for continuous responses into a Bayesian model for multi-response data sets and demonstrates the flexibility of this model through an application to Bayesian additive regression trees (BART), spatial mixed effects (SME) models, and the multivariate spatio-temporal mixed effects model (MSTM).


A joint Bayesian hierarchical model of “multi-type responses” (discrete and continuous responses) and associated Bayesian method are used to analyze bivariate spatial data of daily averaged PM2.5 levels and discrete mortality counts over US counties to address the burden of Bayesian computation for large databases.



Gaussian predictive process models for large spatial data sets.

This work achieves the flexibility to accommodate non‐stationary, non‐Gaussian, possibly multivariate, possibly spatiotemporal processes in the context of large data sets in the form of a computational template encompassing these diverse settings.

Adding Spatially-Correlated Errors Can Mess Up the Fixed Effect You Love

Many statisticians have had the experience of fitting a linear model with uncorrelated errors, then adding a spatially-correlated error term (random effect) and finding that the estimates of the

Improving the performance of predictive process modeling for large datasets

Bayesian Inference for the Spatial Random Effects Model

Spatial statistical analysis of massive amounts of spatial data can be challenging because computation of optimal procedures can break down. The Spatial Random Effects (SRE) model uses a fixed number

Limitations on low rank approximations for covariance matrices of spatial data

Hierarchical Nearest-Neighbor Gaussian Process Models for Large Geostatistical Datasets

A class of highly scalable nearest-neighbor Gaussian process (NNGP) models to provide fully model-based inference for large geostatistical datasets are developed and it is established that the NNGP is a well-defined spatial process providing legitimate finite-dimensional Gaussian densities with sparse precision matrices.

Multivariate spatio-temporal models for high-dimensional areal data with application to Longitudinal Employer-Household Dynamics

This work analyzes Quarterly Workforce Indicators published by the US Census Bureau's Longitudinal Employer-Household Dynamics program and introduces the multivariate spatio-temporal mixed effects model (MSTM), which can be used to efficiently model high-dimensional multivariate spatial areal data sets.

A full scale approximation of covariance functions for large spatial data sets

Summary.  Gaussian process models have been widely used in spatial statistics but face tremendous computational challenges for very large data sets. The model fitting and spatial prediction of such

Spatial Lasso With Applications to GIS Model Selection

This work formalizes this problem as a linear model and proposes the use of Lasso to simultaneously select variables, choose neighborhoods, and estimate parameters, which performs well in numerical examples and is also extended to generalized linear models.

An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach

It is shown that, using an approximate stochastic weak solution to (linear) stochastically partial differential equations, some Gaussian fields in the Matérn class can provide an explicit link, for any triangulation of , between GFs and GMRFs, formulated as a basis function representation.