Pointless Continuous Spatial Surface Reconstruction

The analysis of area-level aggregated summary data is common in many disciplines including epidemiology and the social sciences. Typically, Markov random field spatial models have been employed to acknowledge spatial dependence and allow data-driven smoothing. In this paper, we exploit recent theoretical and computational advances in continuous spatial modeling to carry out the reconstruction of an underlying continuous spatial surface. In particular, we focus on models based on stochastic partial differential equations (SPDEs). We also consider the interesting case in which the aggregate data are supplemented with point data. We carry out Bayesian inference, and in the language of generalized linear mixed models, if the link is linear, an efficient implementation of the model is available via integrated nested Laplace approximations. For nonlinear links, we present two approaches: a fully Bayesian implementation using a Hamiltonian Monte Carlo algorithm, and an empirical Bayes implementation, that is much faster, and is based on Laplace approximations. We examine the properties of the approach using simulation, and then estimate an underlying continuous risk surface for the classic Scottish lip cancer data.

[1]  Y. LindaJ. Combining Incompatible Spatial Data , 2003 .

[2]  Auger‐Methe Marie,et al.  Template Model Builder(TMB)を利用した海洋移動データの空間時間的モデリング , 2017 .

[3]  J. Thorson,et al.  The importance of spatial models for estimating the strength of density dependence. , 2015, Ecology.

[4]  S. Martino Approximate Bayesian Inference for Latent Gaussian Models , 2007 .

[5]  Andrew J. Tatem,et al.  Creating spatial interpolation surfaces with DHS data , 2015 .

[6]  Kiersten B. Johnson,et al.  The DHS wealth index , 2004 .

[7]  Sw. Banerjee,et al.  Hierarchical Modeling and Analysis for Spatial Data , 2003 .

[8]  Jon Wakefield,et al.  Disease mapping and spatial regression with count data. , 2007, Biostatistics.

[9]  Haavard Rue,et al.  Think continuous: Markovian Gaussian models in spatial statistics , 2011, 1110.6796.

[10]  Patrick Brown,et al.  A local-EM algorithm for spatio-temporal disease mapping with aggregated data , 2017 .

[11]  S V Subramanian,et al.  Demographic and health surveys: a profile. , 2012, International journal of epidemiology.

[12]  Alan E Gelfand,et al.  A Spatio-Temporal Downscaler for Output From Numerical Models , 2010, Journal of agricultural, biological, and environmental statistics.

[13]  Jonathan R. Bradley,et al.  Bayesian Spatial Change of Support for Count-Valued Survey Data With Application to the American Community Survey , 2014, 1405.7227.

[14]  K. Mengersen,et al.  A geostatistical model for combined analysis of point-level and area-level data using INLA and SPDE , 2017 .

[15]  Christian P. Robert,et al.  Statistics for Spatio-Temporal Data , 2014 .

[16]  Bob Baulch,et al.  Spatial patterns of poverty in Vietnam and their implications for policy , 2005 .

[17]  Roger Woodard,et al.  Interpolation of Spatial Data: Some Theory for Kriging , 1999, Technometrics.

[18]  J. Besag,et al.  Bayesian image restoration, with two applications in spatial statistics , 1991 .

[19]  P. Brown,et al.  Mapping Cancer Risk in Southwestern Ontario with Changing Census Boundaries , 2012, Biometrics.

[20]  Norbert Henninger,et al.  Spatial determinants of poverty in rural Kenya , 2007, Proceedings of the National Academy of Sciences.

[21]  Patrick Brown,et al.  Local-EM and the EMS Algorithm , 2011 .

[22]  Maurizio Filippone,et al.  A comparative evaluation of stochastic-based inference methods for Gaussian process models , 2013, Machine Learning.

[23]  Montserrat Fuentes,et al.  Model Evaluation and Spatial Interpolation by Bayesian Combination of Observations with Outputs from Numerical Models , 2005, Biometrics.

[24]  Radford M. Neal MCMC Using Hamiltonian Dynamics , 2011, 1206.1901.

[25]  Kenya,et al.  Kenya Demographic and Health Survey 2014 , 2015 .

[26]  Patrick Brown,et al.  Log Gaussian Cox processes and spatially aggregated disease incidence data , 2012, Statistical methods in medical research.

[27]  RockOn Team,et al.  Re: Attenuation compensation in single-photon emission tomography: a comparative evaluation. , 1983, Journal of nuclear medicine : official publication, Society of Nuclear Medicine.

[28]  Peter J. Diggle,et al.  Spatial and spatio-temporal Log-Gaussian Cox processes:extending the geostatistical paradigm , 2013, 1312.6536.

[29]  Finn Lindgren,et al.  Bayesian Spatial Modelling with R-INLA , 2015 .