Some recent work on multivariate Gaussian Markov random fields

Some recent work on conditional formulation of multivariate Gaussian Markov random fields is presented. The focus is on model constructions by compatible conditionals and coregionalization. Special attention is given to multivariate generalizations of univariate models. Beginning with univariate model constructions, a survey of key approaches to formulating multivariate extensions is presented. Two challenges in the formulation and implementation of multivariate models are highlighted: (1) entanglement of spatial and non-spatial components, and (2) enforcement for positivity condition. Managing the two challenges by decomposition, separation, and constrained parameterization is discussed. Also highlighted is the challenge of flexible modeling of (conditional) cross-spatial dependencies and, in particular, asymmetric cross-spatial dependencies. Interpretation of asymmetric cross-spatial dependencies is also discussed. A coregionalization framework which connects and unifies the various lines of model development is presented. The framework enables a systematic development of a broad range of models via linear and spatially varying coregionalization, respectively, with extensions to locally adaptive models. Formulation of multivariate models over variable-specific lattices is discussed. Selected models are illustrated with examples of Bayesian multivariate and spatiotemporal disease mapping. Potential applications of coregionalization models in imaging analysis, covariance modeling, dimension reduction, and latent variable analysis are briefly mentioned.

[1]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[2]  J. Andrew Royle,et al.  A hierarchical approach to multivariate spatial modeling and prediction , 1999 .

[3]  M. Daniels,et al.  Conditionally Specified Space-Time Models for Multivariate Processes , 2006 .

[4]  Gilbert MacKenzie,et al.  On modelling mean‐covariance structures in longitudinal studies , 2003 .

[5]  H. Rue,et al.  An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach , 2011 .

[6]  Ying C MacNab,et al.  On Bayesian shared component disease mapping and ecological regression with errors in covariates , 2010, Statistics in medicine.

[7]  Ruth Salway,et al.  A statistical framework for ecological and aggregate studies , 2001 .

[8]  M. Genton.  Separable approximations of space‐time covariance matrices , 2007 .

[9]  Duncan Lee,et al.  A Bayesian localized conditional autoregressive model for estimating the health effects of air pollution , 2013, Biometrics.

[10]  A. Gelfand,et al.  Proper multivariate conditional autoregressive models for spatial data analysis. , 2003, Biostatistics.

[11]  C. F. Sirmans,et al.  Nonstationary multivariate process modeling through spatially varying coregionalization , 2004 .

[12]  M. D. Ugarte,et al.  One-dimensional, two-dimensional, and three dimensional B-splines to specify space –time interactions in Bayesian disease mapping: Model fitting and model identifiability , 2017 .

[13]  James S Hodges,et al.  Generalized spatial structural equation models. , 2005, Biostatistics.

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

[15]  Ying C MacNab,et al.  On Gaussian Markov random fields and Bayesian disease mapping , 2011, Statistical methods in medical research.

[16]  Mark J. Brewer,et al.  Variable smoothing in Bayesian intrinsic autoregressions , 2007 .

[17]  M. Martínez-Beneito.  A general modelling framework for multivariate disease mapping , 2013 .

[18]  Peter Congdon,et al.  A spatially adaptive conditional autoregressive prior for area health data , 2008 .

[19]  Victor De Oliveira,et al.  Bayesian analysis of conditional autoregressive models , 2012 .

[20]  Nicola G. Best,et al.  A shared component model for detecting joint and selective clustering of two diseases , 2001 .

[21]  A. Dawid.  Some matrix-variate distribution theory: Notational considerations and a Bayesian application , 1981 .

[22]  J. Besag.  Spatial Interaction and the Statistical Analysis of Lattice Systems , 1974 .

[23]  Mike Rees,et al.  5. Statistics for Spatial Data , 1993 .

[24]  Tony Pourmohamad,et al.  Multivariate Stochastic Process Models for Correlated Responses of Mixed Type , 2016 .

[25]  L. Held,et al.  Towards joint disease mapping , 2005, Statistical methods in medical research.

[26]  Hans Wackernagel,et al.  Multivariate Geostatistics: An Introduction with Applications , 1996 .

[27]  P Botella-Rocamora,et al.  A unifying modeling framework for highly multivariate disease mapping , 2015, Statistics in medicine.

[28]  T. Goicoa,et al.  In spatio-temporal disease mapping models, identifiability constraints affect PQL and INLA results , 2018, Stochastic Environmental Research and Risk Assessment.

[29]  J. Besag,et al.  On conditional and intrinsic autoregressions , 1995 .

[30]  J. Zidek,et al.  Multivariate spatial interpolation and exposure to air pollutants , 1994 .

[31]  Sumio Watanabe,et al.  Asymptotic Equivalence of Bayes Cross Validation and Widely Applicable Information Criterion in Singular Learning Theory , 2010, J. Mach. Learn. Res..

[32]  Noel A Cressie,et al.  A spatial model for multivariate lattice data , 2007 .

[33]  R. Kass,et al.  Nonconjugate Bayesian Estimation of Covariance Matrices and its Use in Hierarchical Models , 1999 .

[34]  Hao Zhang,et al.  Maximum‐likelihood estimation for multivariate spatial linear coregionalization models , 2007 .

[35]  Ying C MacNab.  MacNab spatial regression models − On identification in Bayesian disease mapping and ecological , 2014 .

[36]  Alan E. Gelfand,et al.  Model choice: A minimum posterior predictive loss approach , 1998, AISTATS.

[37]  Robert J. Plemmons,et al.  Nonnegative Matrices in the Mathematical Sciences , 1979, Classics in Applied Mathematics.

[38]  A. O'Hagan,et al.  Bayesian emulation of complex multi-output and dynamic computer models , 2010 .

[39]  J. Hodges,et al.  Adding Spatially-Correlated Errors Can Mess Up the Fixed Effect You Love , 2010 .

[40]  Bradley P Carlin,et al.  Generalized Hierarchical Multivariate CAR Models for Areal Data , 2005, Biometrics.

[41]  Ying C. MacNab,et al.  Spline smoothing in Bayesian disease mapping , 2007 .

[42]  L. Fahrmeir,et al.  Adaptive Gaussian Markov random fields with applications in human brain mapping , 2007 .

[43]  M. Pourahmadi.  Maximum likelihood estimation of generalised linear models for multivariate normal covariance matrix , 2000 .

[44]  Mohsen Pourahmadi,et al.  High-Dimensional Covariance Estimation , 2013 .

[45]  Mohsen Pourahmadi,et al.  High-Dimensional Covariance Estimation , 2013 .

[46]  Y. MacNab.  Bayesian multivariate disease mapping and ecological regression with errors in covariates: Bayesian estimation of DALYs and ‘preventable’ DALYs , 2009, Statistics in medicine.

[47]  Andrew B. Lawson,et al.  Bayesian Disease Mapping: Hierarchical Modeling in Spatial Epidemiology , 2008 .

[48]  M. Pourahmadi.  Covariance Estimation: The GLM and Regularization Perspectives , 2011, 1202.1661.

[49]  Stephan R. Sain,et al.  spam: A Sparse Matrix R Package with Emphasis on MCMC Methods for Gaussian Markov Random Fields , 2010 .

[50]  Peter Congdon.  A spatial structural equation model for health outcomes , 2008 .

[51]  P. Speckman,et al.  Posterior distribution of hierarchical models using CAR(1) distributions , 1999 .

[52]  K. Mardia.  Multi-dimensional multivariate Gaussian Markov random fields with application to image processing , 1988 .

[53]  Noel A Cressie,et al.  Statistics for Spatio-Temporal Data , 2011 .

[54]  R. Varga,et al.  Block diagonally dominant matrices and generalizations of the Gerschgorin circle theorem , 1962 .

[55]  Ying C MacNab,et al.  Regression B‐spline smoothing in Bayesian disease mapping: with an application to patient safety surveillance , 2007, Statistics in medicine.

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

[57]  Leonhard Held,et al.  Gaussian Markov Random Fields: Theory and Applications , 2005 .

[58]  Duncan Lee,et al.  A comparison of conditional autoregressive models used in Bayesian disease mapping. , 2011, Spatial and spatio-temporal epidemiology.

[59]  Assessment of a bayesian multivariate interpolation approach for health impact studies , 1998 .

[60]  A. Gelfand,et al.  A Bayesian coregionalization approach for multivariate pollutant data , 2003 .

[61]  V. Zadnik,et al.  Effects of Residual Smoothing on the Posterior of the Fixed Effects in Disease‐Mapping Models , 2006, Biometrics.

[62]  T. Simons,et al.  Spatial autocorrelation and autoregressive models in ecology , 2002 .

[63]  Ying C MacNab,et al.  An innovative application of Bayesian disease mapping methods to patient safety research: a Canadian adverse medical event study , 2006, Statistics in medicine.

[64]  L Knorr-Held,et al.  Bayesian modelling of inseparable space-time variation in disease risk. , 2000, Statistics in medicine.

[65]  Sudipto Banerjee,et al.  Towards a Multidimensional Approach to Bayesian Disease Mapping. , 2017, Bayesian analysis.

[66]  James S Hodges,et al.  Modeling Longitudinal Spatial Periodontal Data: A Spatially Adaptive Model with Tools for Specifying Priors and Checking Fit , 2008, Biometrics.

[67]  Marc G. Genton,et al.  Cross-Covariance Functions for Multivariate Geostatistics , 2015, 1507.08017.

[68]  A. Raftery,et al.  Strictly Proper Scoring Rules, Prediction, and Estimation , 2007 .

[69]  Rama Chellappa,et al.  Estimation and choice of neighbors in spatial-interaction models of images , 1983, IEEE Trans. Inf. Theory.

[70]  Ying C MacNab,et al.  Linear models of coregionalization for multivariate lattice data: a general framework for coregionalized multivariate CAR models , 2016, Statistics in medicine.

[71]  Dongchu Sun,et al.  A Bivariate Bayes Method for Improving the Estimates of Mortality Rates With a Twofold Conditional Autoregressive Model , 2001 .

[72]  D. Harville.  Matrix Algebra From a Statistician's Perspective , 1998 .

[73]  Norman E. Breslow,et al.  Estimation of Disease Rates in Small Areas: A new Mixed Model for Spatial Dependence , 2000 .

[74]  M. Pourahmadi.  Joint mean-covariance models with applications to longitudinal data: Unconstrained parameterisation , 1999 .

[75]  Peter D. Hoff,et al.  Separable covariance arrays via the Tucker product, with applications to multivariate relational data , 2010, 1008.2169.

[76]  Jeffrey S. Morris,et al.  Functional CAR Models for Large Spatially Correlated Functional Datasets , 2016, Journal of the American Statistical Association.

[77]  M. Pourahmadi.  Cholesky Decompositions and Estimation of A Covariance Matrix: Orthogonality of Variance--Correlation Parameters , 2007 .

[78]  Bradley P. Carlin,et al.  Bayesian measures of model complexity and fit , 2002 .

[79]  Ying C MacNab,et al.  Linear models of coregionalization for multivariate lattice data: Order-dependent and order-free cMCARs , 2016, Statistical methods in medical research.

[80]  Xiao-Li Meng,et al.  Modeling covariance matrices in terms of standard deviations and correlations, with application to shrinkage , 2000 .

[81]  Carlo Trivisano,et al.  A multivariate CAR model for improving the estimation of relative risks. , 2009, Statistics in medicine.

[82]  Stephan R. Sain,et al.  A spatial analysis of multivariate output from regional climate models , 2011, 1104.2703.

[83]  G. MacKenzie,et al.  Modelling conditional covariance in the linear mixed model , 2007 .

[84]  Bradley P Carlin,et al.  Order‐free co‐regionalized areal data models with application to multiple‐disease mapping , 2007, Journal of the Royal Statistical Society. Series B, Statistical methodology.