Modelling Nonstationary Spatial Lag Models with Hidden Markov Random Fields

Abstract One of the basic assumptions in spatial statistic is second-order stationarity, which implies homogeneity and isotropy. However, when using a spatial random field framework to model socio-economical or epidemiological data - just to mention two examples - it is often unreasonable to believe that the relationship between variables could be modelled as a realization of a unique stationary process. In order to provide a more realistic representation, we introduce a latent process which drives the value of the coefficients in a Cliff-Ord-type spatial autoregressive linear model identifying groups of observations with a similar behaviour. The latent process evolves as a Hidden Markov Random Field. This structure allows the topology of the problem to be taken into account when identifying groups. A simulation exercise is performed to investigate the influence of parameter values - estimated via a Markov chain Monte Carlo procedure - on the accuracy of the results. Criteria to perform model comparison in order to establish the optimal number of clusters are also provided. A case study referred to hedonic house prices in Boston illustrates the advantages of the proposed modelling strategy.

[1]  Dennis M. Gorman,et al.  Quantifying geographic variations in associations between alcohol distribution and violence: a comparison of geographically weighted regression and spatially varying coefficient models , 2007 .

[2]  A. Gelman.  Prior distributions for variance parameters in hierarchical models (comment on article by Browne and Draper) , 2004 .

[3]  C. F. Sirmans,et al.  Spatial Modeling With Spatially Varying Coefficient Processes , 2003 .

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

[5]  M. Charlton,et al.  Geographically Weighted Regression: A Natural Evolution of the Expansion Method for Spatial Data Analysis , 1998 .

[6]  P. Green,et al.  Hidden Markov Models and Disease Mapping , 2002 .

[7]  Alan Edelman,et al.  Julia: A Fresh Approach to Numerical Computing , 2014, SIAM Rev..

[8]  Brian J. Smith,et al.  Convergence diagnostics for MCMC draws of a categorical variable , 2017, 1706.04919.

[9]  Martin Charlton,et al.  Geographically weighted discriminant analysis , 2007 .

[10]  Hui Wang,et al.  A Segmentation Approach for Stochastic Geological Modeling Using Hidden Markov Random Fields , 2017, Mathematical Geosciences.

[11]  George Christakos,et al.  Random Field Models in Earth Sciences , 1992 .

[12]  W. Cleveland,et al.  Locally Weighted Regression: An Approach to Regression Analysis by Local Fitting , 1988 .

[13]  J. Besag.  On the Statistical Analysis of Dirty Pictures , 1986 .

[14]  Bradley P. Carlin,et al.  Bayesian areal wombling for geographical boundary analysis , 2005 .

[15]  Daniel A. Griffith,et al.  Spatial-Filtering-Based Contributions to a Critique of Geographically Weighted Regression (GWR) , 2008 .

[16]  Loay Kadom Abood,et al.  Detecting brain tumor in Magnetic Resonance Images using Hidden Markov Random Fields and Threshold techniques , 2014, 2014 IEEE Student Conference on Research and Development.

[17]  H. Omre,et al.  Impact of rock-physics depth trends and Markov random fields on hierarchical Bayesian lithology/fluid prediction , 2010 .

[18]  Jurg Ott,et al.  Scan statistics in human gene mapping. , 2012, American journal of human genetics.

[19]  Daniel A. Griffith,et al.  Some robustness assessments of Moran eigenvector spatial filtering , 2017 .

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

[21]  Duncan Lee,et al.  CARBayes: An R Package for Bayesian Spatial Modeling with Conditional Autoregressive Priors , 2013 .

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

[23]  Fitting spatial regressions to large datasets using unilateral approximations , 2018 .

[24]  Peter A. Rogerson,et al.  A Statistical Method for the Detection of Geographic Clustering , 2010 .

[25]  Peter J. Diggle,et al.  Statistical analysis of spatial point patterns , 1983 .

[26]  A. Getis,et al.  Using AMOEBA to Create a Spatial Weights Matrix and Identify Spatial Clusters , 2006 .

[27]  Mark J. Embrechts,et al.  On the Use of the Adjusted Rand Index as a Metric for Evaluating Supervised Classification , 2009, ICANN.

[28]  J. Kaashoek,et al.  Potentialised Partial Differential Equations in Economic Geography and Spatial Economics: Multiple Dimensions and Control , 1998 .

[29]  Florence Tupin,et al.  Unsupervised classification of radar images using hidden Markov chains and hidden Markov random fields , 2003, IEEE Trans. Geosci. Remote. Sens..

[30]  J. LeSage,et al.  Spatial Econometric Models , 2010 .

[31]  W. Zucchini,et al.  Hidden Markov Models for Time Series: An Introduction Using R , 2009 .

[32]  D. Rubinfeld,et al.  Hedonic housing prices and the demand for clean air , 1978 .

[33]  Huiyan Sang,et al.  Spatial Homogeneity Pursuit of Regression Coefficients for Large Datasets , 2019, Journal of the American Statistical Association.

[34]  Minfeng Deng,et al.  An Anisotropic Model for Spatial Processes , 2007 .

[35]  Giuseppe Arbia,et al.  A primer for Spatial Econometrics , 2014 .

[36]  Nial Friel,et al.  Spatial hidden Markov models and species distributions , 2018 .

[37]  Stephen M. Smith,et al.  Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm , 2001, IEEE Transactions on Medical Imaging.

[38]  Nathalie Peyrard,et al.  Classification method for disease risk mapping based on discrete hidden Markov random fields. , 2012, Biostatistics.

[39]  T. Raa.  Spatial economics: density, potential, and flow: Martin BECKMANN and Tönu PUU Volume 14 in: Studies in Regional Science and Urban Economics, North-Holland, Amsterdam, 1985, xii + 276 pages, Dfl. 125.00 , 1986 .

[40]  P. Krugman.  Geography and Trade , 1992 .

[41]  John F. McDonald,et al.  Locally Weighted Maximum Likelihood Estimation: Monte Carlo Evidence and an Application , 2004 .

[42]  Charles J. Geyer,et al.  EXTENDING PSEUDO-LIKELIHOOD FOR POTTS MODELS , 2011 .

[43]  M. Kulldorff.  A spatial scan statistic , 1997 .

[44]  Ajay Jasra,et al.  Markov Chain Monte Carlo Methods and the Label Switching Problem in Bayesian Mixture Modeling , 2005 .

[45]  Pedram Ghamisi,et al.  Spectral–Spatial Classification of Hyperspectral Images Based on Hidden Markov Random Fields , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[46]  I. Epifani,et al.  On the Population Density Distribution Across Space: A Probabilistic Approach , 2013 .

[47]  Peter Congdon,et al.  Bayesian model choice based on Monte Carlo estimates of posterior model probabilities , 2006, Comput. Stat. Data Anal..

[48]  W. H. Womble,et al.  Differential systematics. , 1951, Science.

[49]  J. Kaashoek,et al.  On potentialized partial differential equations in theoretical spatial economics , 1994 .

[50]  A. Venables,et al.  The Spatial Economy: Cities, Regions, and International Trade , 1999 .

[51]  A. Stewart Fotheringham,et al.  Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity , 2010 .

[52]  P. Whittle.  ON STATIONARY PROCESSES IN THE PLANE , 1954 .

[53]  A frequency domain test for isotropy in spatial data models , 2017 .

[54]  G. Arbia.  Spatial Econometrics , 2006, Encyclopedia of Big Data.

[55]  G. Arbia.  On second-order non-stationarity in two dimensional lattice processes , 1989 .

[56]  Steven Farber,et al.  Moving Window Approaches for Hedonic Price Estimation: An Empirical Comparison of Modelling Techniques , 2008 .

[57]  Walter Christaller.  Die zentralen Orte in Süddeutschland , 1980 .

[58]  Martin Charlton,et al.  Spatial Nonstationarity and Autoregressive Models , 1998 .

[59]  S. Ait-Aoudia,et al.  Medical Image Segmentation Using Particle Swarm Optimization , 2014, 2014 18th International Conference on Information Visualisation.