Abstract
In this paper we propose a blockwise Euclidean likelihood method for the estimation of a spatial binary field obtained by thresholding a latent Gaussian random field. The moment conditions used in the Euclidean likelihood estimator derive from the score of the composite likelihood based on marginal pairs. A feature of this approach is that it is possible to obtain computational benefits with respect to the pairwise likelihood depending on the choice of the spatial blocks. A simulation study and an analysis on cancer mortality data compares the two methods in terms of statistical and computational efficiency. We also study the asymptotic properties of the proposed estimator.
Similar content being viewed by others
References
Adler J (2008) Some new random field tools for spatial analysis. Stoch Environ Res Risk Assess 22:809–822
Albert P, McShane L (1995) A generalized estimating equations approach for spatially correlated binary data: applications to the analysis of neuroimaging data. Biometrics 51:627–638
Bevilacqua M, Gaetan C (2014) Comparing composite likelihood methods based on pairs for spatial gaussian random fields. Statistics and Computing, to appear.
Bevilacqua M, Gaetan C, Mateu J, Porcu E (2012) Estimating space and space-time covariance functions: a weighted composite likelihood approach. J Am Stat Assoc Theory Methods 107:268–280
Caragea P, Smith R (2006) Approximate likelihoods for spatial processes (Tech. Ref.). Department of Statistics, Iowa State University
Cressie N (1993) Statistics for spatial data (revised ed). Wiley, New York
Cressie N, Johannesson G (2008) Fixed rank kriging for very large spatial data sets. J Royal Stat Soc Ser B 70:209–226
Eidsvik J, Shaby B, Reich B, Wheeler M, Niemi J (2013). Estimation and prediction with block composite likelihoods using parallel computing. J Comput Graph Stat 23:295–315
Fuentes M (2007) Approximate likelihood for large irregularly spaced spatial data. J Am Stat Assoc 102:321–331
Gneiting T, Kleiber W, Schlather M (2010) Matérn cross-covariance functions for multivariate random fields. J Am Stat Assoc 105:1167–1177
Heagerty P, Lele S (1998) A composite likelihood approach to binary spatial data. J Am Stat Assoc 93:1099–1111
Heagerty P, Zeger SL (1996) Lorelogram: a regression approach to exploring dependence in longitudinal categorical responses. J Am Stat Assoc Theory Methods 93:150–162
Jenish N, Prucha I (2009) Central limit theorems and uniform laws of large numbers for arrays of random fields. J Econom 150:86–98
Journel A (1983) Nonparametric estimation of spatial distributions. J Int Assoc Math Geol 15:468–485
Kitamura Y (1997) Empirical likelihood methods with weakly dependent processes. Ann Stat 25:2084–2102
Kleiber W, Porcu E (2014) Nonstationary matrix covariances: compact support, long range dependence and quasi-arithmetic constructions. Stoch Environ Res Risk Assess, (to appear)
Kovitz JL, Christakos G (2004) Spatial statistics of clustered data. Stoch Environ Res Risk Assess 18:147–166
Lahiri SN (1999) Theoretical comparisons of block bootstrap methods. Ann Stat 27:386–404
Lee Y, Lahiri S (2002) Least squares variogram fitting by spatial subsampling. J Royal Stat Soc B 64:837–854
Lin L, Zhang R (2001) Blockwise Euclidean likelihood for weakly dependent processes. Stat Probab Lett 53:143–152
Lin P, Clayton K (2005) Analysis of binary spatial data by quasi-likelihood estimating equations. Ann Stat 33:542–555
Lindsay B (1988) Composite likelihood methods. Contemp Math 80:221–239
Lunardon N, Pauli F, Ventura L (2012) A note on empirical likelihoods derived from pairwise score functions. J Stat Comput Simul 8:1405–1414
Martinez-Ruiz F, Mateu J, Montes F, Porcu E (2010) Mortality risk assessment through stationary space-time covariance functions. Stoch Environ Res Risk Assess 24:519–526
Matheron G (1976) A simple substitute for the conditional expectation: the disjuntive kriging. In: Guarascio M, David M, Huijbrechts C (eds) Advanced geostatistics in the mining industry. Springer, Dordrecht, pp 221–236
Newey W, Smith RJ (2004) Higher order properties of GMM and generalized empirical likelihood estimators. Econometrica 72:219–255
Nordman D (2008) A blockwise empirical likelihood for spatial data. Statistica Sinica 18:1111–1129
Nordman D, Caragea P (2008) Point and interval estimation of variogram models using spatial empirical likelihood. J Am Stat Assoc 103:350–361
Owen AB (1988) Empirical likelihood ratio confidence intervals for a single functional. Biometrika 75:237–249
Owen AB (1990) Empirical likelihood confidence regions. Ann Stat 18:90–120
Owen AB (1991) Empirical likelihood for linear models. Ann Stat 19:1725–1747
Owen AB (2001) Empirical likelihood. Chapman & Hall/CRC, London
Pace L, Salvan A, Sartori N (2011) Adjusting composite likelihood ratio statistics. Statistica Sinica 21:129–148
Padoan SA, Ribatet M, Sisson SA (2010) Likelihood-based inference for max-stable processes. J Am Stat Assoc Theory Methods 105:263–277
Porcu E, Daley DJ, Buhmann M, Bevilacqua M (2013) Radial basis functions with compact support for multivariate geostatistics. Stoch Environ Res Risk Assess 27:909–922
Qin J, Lawless J (1994) Empirical likelihood and general estimating equations. Ann Stat 22:300–325
Rue H, Tjelmeland H (2002) Fitting Gaussian Markov random fields to Gaussian fields. Scandinavian J Stat 29:31–49
Ruiz-Medina MD, Angulo JM, Anh VV (2008) Multifractality in space-time statistical models. Stoch Environ Res Risk Assess 22:81–86
Sherman M (1996) Variance estimation for statistics computed from spatial lattice data. J Royal Stat Soc B 58:509–523
Sherman M, Apanasovich T, Carroll R (2006) On estimation of binary autologistic spatial models. J Stat Comput Simul 76:167–179
Sherman M, Carlstein E (1994) Nonparametric estimation of the moments of a general statistic computed from spatial data. J Am Stat Assoc Theory Methods 89:496–500
Stein M (2008) A modeling approach for large spatial datasets. J Korean Stat Soc 37:3–10
Stein M, Chi Z, Welty L (2004) Approximating likelihoods for large spatial data sets. J Royal Stat Soc B 66:275–296
Sun Y, Li B, Genton MG (2012) Geostastistics for large space-time datasets. In: Porcu E, Montero JM, Schlather M (eds) Space-time processes and challenges related to environmental problems
Varin C, Reid N, Firth D (2011) An overview of composite likelihood methods. Statistica Sinica 21:5–42
Vecchia A (1988) Estimation and model identification for continuous spatial processes. J Royal Stat Soc B 50:297–312
Yu H-L, Kovolos A, Christakos G, Chen J-C, Warmerdam S, Dev B (2007) Interactive spatiotemporal modelling of health systems: the SEKS-GUI framework. Stoch Environ Res Risk Assess 21:555–572
Acknowledgments
Research work of Moreno Bevilacqua was partially supported by grant FONDECYT 11121408 from the Chilean government. Research work of Emilio Porcu is supported by Proyecto Fondecyt Regular N. 1130647 from the Chilean Government. We are indebted with Prof. Daniel Nordman for having provided us with the cancer mortality data.
Author information
Authors and Affiliations
Corresponding author
Appendix
Appendix
The appendix collects the proofs of the asymptotic results. First of all, we introduce the following notation: \(\nabla _{\varvec{\theta }}\) and \(\nabla _{\varvec{\lambda }}\) are the first derivative operators for \({\varvec{\theta }}\) and \({\varvec{\lambda }}\) respectively, while \(\nabla _{{\varvec{\theta }}{\varvec{\theta }}}\), \(\nabla _{{\varvec{\lambda }}{\varvec{\lambda }}}\) and \(\nabla _{{\varvec{\theta }}{\varvec{\lambda }}}\) indicate second and cross derivatives and are defined accordingly. Similarly, for a certain function \(\widehat{R}({\varvec{\theta }},{\varvec{\lambda }})\) defined below, \(\widehat{R}_{\varvec{\theta }}({\varvec{\theta }},{\varvec{\lambda }})\) is its first derivative with respect to \({\varvec{\theta }}\). Derivatives with respect to \({\varvec{\lambda }}\), second derivatives and cross derivatives are defined in a similar manner. In addition to that, we use
instead of the expression in Eq. (20). As noticed in Newey and Smith (2004) this choice does not change the estimator but it simplifies the asymptotic analysis.
Proof
(Proof of Theorem 1) We have to show that, for some \(\delta >0\), \(P(\Vert \widehat{\varvec{\theta }}-{\varvec{\theta }}_0\Vert >\delta )\rightarrow 0\) as \(n\rightarrow \infty \). By continuity of \(Q({\varvec{\theta }})\) and the assumption that \({\varvec{\theta }}_0\) is the unique minimizer, we have that, for some \(\varepsilon >0\), \(\{ \Vert \widehat{\varvec{\theta}}-{\varvec{\theta}}_0\Vert > \delta \} \Longrightarrow \{ \vert \ Q(\widehat{\varvec{\theta }})-Q({\varvec{\theta}}_0)\vert >\varepsilon \} \). This is, the latter set contains the former. Hence, \(P(\Vert \widehat{\varvec{\theta }}-{\varvec{\theta }}_0\Vert >\delta )\le P(\vert \ Q(\widehat{\varvec{\theta }})-Q({\varvec{\theta }}_0)\vert >\varepsilon )\). By assumptions C1–C2 and arguments similar to those in Theorem 2 we have the following uniform convergence condition
Therefore,
where the latter inequality follows from the triangular inequality and the uniform convergence condition in (23). This implies \(P(\Vert \widehat{\varvec{\theta }}-{\varvec{\theta }}_0\Vert >\delta )\le P(\vert \ Q(\widehat{\varvec{\theta }})-Q({\varvec{\theta }}_0)\vert >\varepsilon )\rightarrow 0\) as \(n\rightarrow \infty \). Hence, \(\widehat{\varvec{\theta }}\rightarrow _p{\varvec{\theta }}_0\). Before showing asymptotic normality we show that the estimate of the Lagrange multiplier \(\widehat{\varvec{\lambda }}\) converges to zero in probability. Similarly to (7) we have that
By a mean value argument, assumptions C1-C2, results in Theorem 2 and continuous mapping theorem we get
In analogy with (5) let us define
and
Then by simple calculations we can write
The first order conditions of \(\widehat{R}(\widehat{\varvec{\theta }},\widehat{\varvec{\lambda }})\) with respect to \({\varvec{\theta }}\) and \({\varvec{\lambda }}\) are
Let us now take a mean value expansion of the first order conditions (25) and (26) about the true values \((\varvec{\theta }^\intercal ,{\varvec{\lambda }}^\intercal )^\intercal =({\varvec{\theta }}_0^\intercal ,{\bf{0}}^\intercal )^\intercal \)
More compactly,
By the UWLLN \(\frac{1}{b^d}\widehat{R}_{\varvec{\theta }{\varvec{\theta }}}(\dot{\varvec{\theta }},\dot{\varvec{\lambda }})\rightarrow _p{\bf{0}}\), \(b^d\widehat{R}_{\varvec{\lambda }{\varvec{\lambda }}}(\dot{\varvec{\theta }},\dot{\varvec{\lambda }})\rightarrow _p{\varvec{\varSigma }}({\varvec{\theta }}_0)\) and \(\widehat{R}_{\varvec{\lambda }{\varvec{\theta }}}(\dot{\varvec{\theta }},\dot{\varvec{\lambda }})\rightarrow _p-\nabla _{\varvec{\theta }}{\varvec{m}}({\varvec{\theta }}_0)\). Hence,
where \({\varvec{\varOmega }}({\varvec{\theta }}_0)=(\nabla _{\varvec{\theta }}{\varvec{m}}({\varvec{\theta }}_0)^\intercal {\varvec{\varSigma }}({\varvec{\theta }}_0)^{-1}\nabla _{\varvec{\theta }}{\varvec{m}}({\varvec{\theta }}_0))^{-1}\) and \({\varvec{\varLambda }}({\varvec{\theta }}_0)={\varvec{\varSigma }}({\varvec{\theta }}_0)^{-1}-{\varvec{\varSigma }}({\varvec{\theta }}_0)^{-1}\nabla _{\varvec{\theta }}{\varvec{m}}({\varvec{\theta }}_0){\varvec{\varOmega }}({\varvec{\theta }}_0)\nabla _{\varvec{\theta }}{\varvec{m}}({\varvec{\theta }}_0)^\intercal {\varvec{\varSigma }}({\varvec{\theta }}_0)^{-1}\). The result follows from an application of the CLT and the continuous mapping theorem. \(\square \)
Proof
(Proof of Theorem 2) Assume \(\widehat{\varvec{\theta }}\) consistent and define the following mean value expansion around \({\varvec{\theta }}_0\)
By rearranging, taking the Euclidean norm and the Cauchy-Schwarz inequality we get
The last line follows from the consistency of \(\widehat{\varvec{\theta }}\) and from the fact that \(b^d=o(\sqrt{n})\). Thus,
The result follows from the application of the triangular inequality and UWLLN. \(\square \)
Rights and permissions
About this article
Cite this article
Bevilacqua, M., Crudu, F. & Porcu, E. Combining Euclidean and composite likelihood for binary spatial data estimation. Stoch Environ Res Risk Assess 29, 335–346 (2015). https://doi.org/10.1007/s00477-014-0938-8
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00477-014-0938-8