Bayesian inference applied to linear regression model and spatial model: An approach on the covariance structure between geostatistical data




Spatial Statistics; Geostatistics; Bayesian inference; Markov chains Monte Carlo; DIC; Linear Regression Model; Mean square error.


Statistical models serve to describe the probabilistic behavior of phenomena of interest, allowing them to be analyzed, predicted and made relevant decisions. Linear regression models are widely used in several areas. These models have strong assumptions such as independence between errors that in general do not fit spatial data, since these data allow dependence on the error covariance structure. Therefore, linear regression models can be compared with spatial models. Spatial data can be divided into 3 types: dot pattern, area data and geostatistical data. This work aims to evaluate linear regression models initially and later to compare them to spatial models for geostatistical data through the exponential covariance function. Unknown parameters are found in these models and the inference adopted in this work is Bayesian for allowing the expert's initial belief to be incorporated into the modeling, increasing the amount of information evaluated and therefore improving the estimates. When adjusting the models under simulated data sets, it is possible to verify the ability of the adjustments to recover the true values of the parameters and select the true model. This article is the result of an interest in analyzing the adjustment of the linear regression model with a set of artificial data with spatial dependence and comparing this to the adjustment of the spatial model, more specifically, based on geostatistical data.


Bernardo, J. M., & Smith, A. F. M. (1994). Bayesian theory(1st ed.). Wiley.

Borgoni, R., & Billari, F. C. (2003). Bayesian spatial analysis of demographic survey data: An application to contraceptive use at first sexual intercourse. Demographic Research,8, 61-92.

Casella, G., & Berger, R. (2002). Statistical inference. Thomson Learning.

Cressie, N. (1993). Statistics for spatial data. J. Wiley.

Câmara, G., & Ortiz, M. (1998). Sistemas de informações geográficas para aplicações ambientais e cadastrais: uma visão geral.In: CONGRESSOBRASILEIRO DE ENGENHARIA AGRÍCOLA.,27, 59-82.

Dawn B. Woodard, R. L. W., & O’Connell, M. A. (2010). Spatial inference of nitrate concentrations in groundwater. Journal of Agricultural, Biological, and Environmental Statistics,15(2), 209-227.

Ehlers, R. S. (2003). Introdução a inferência bayesiana [Computer software manual]. Curitiba. Retrieved from<>

Gamerman, D. (2004). Markov chain monte carlo: Stochastic simulation for bayesian inference (1st ed.). Chapman and Hall/CRC.

Gamerman, D., & Lopes, H. F. (2006). Markov chain monte carlo: Stochastics imulation for bayesian inference (2nd ed.). Chapman and Hall/CRC.

Gelfand, A. E., & Smith, A. F. M. (1990). amping-based approaches to calculating marginal densities. Journal of the American Statistical Association,85(410),398–409.

Geman, S., & Geman, D. (1984). Stochastic relaxation, gibbs distributions, andthe bayesian restoration of images. IEEE Transactions on pattern analysis and machine intelligence,57(6), 721–741.

Hastings, W. (1970). Monte carlo sampling methods using markov chains and their applications. Biometrika,57, 97–109.

Krige, D. (1951). A statistical approaches to some basic mine valuation problemson the witwatersrand. Journal of the Chemical, Metallurgical and Mining Societyof South Africa,52, 119-139.

Macera, M. H. C. (2011). Uso dos métodos clássico e bayesiano para os modelosnão-lineares heterocedásticos simétricos (Ciências de Computação e MatemáticaComputacional). Universidade de São Paulo, São Carlos.

Metropolis, N. (1953). Equation of state calculations by fast computing machines. The journal of chemical physics,21(6), 1087–1092.

Migon, H., Gamerman, D., & Louzada, F. (2014). Statistical inference: An integrated approach, second edition. Taylor & Francis.

Morettin, P. A., & de O. Bussab, W. (2010). Estatistica básica(3rd ed.). Editora Saraiva.

O’Hagan, A., & Kendall. (1994). Bayesian inference(1st ed.). Edward Arnold.

Paulino, C. D., Amaral Turkman, M. A., Murteira, B., & Silva, G. L. (2018). Estatística bayesiana(2a ed.). Lisboa: Fundação Calouste Gulbenkian.

Pereira A.S. et al. (2018). Metodologia da pesquisa científica. [e-book]. Santa Maria. Ed. UAB/NTE/UFSM.

R Core Team. (2019). R: A language and environment for statistical computing [Computer software manual]. Vienna, Austria. Retrieved from<>

Robert, C. P., & Casella, G. (2004). Monte carlo statistical methods. Secaucus, NJ, USA: Springer-Verlag New York, Inc.

Roberts, G. O., Gelman, A., & Gilks, W. R. (1997). Weak convergence and optimal scaling of random walk metropolis algorithms. The Annals of Applied Probability,7, 110-120.

Roberts, G. O., & Rosenthal, J. S. (2001). Optimal scaling for various metropolis-hastings algorithms. Statistical Science,16, 351-367.

Smith, R. L. (1996). Estimating nonstationary spatial correlations. Relatório técnico, Cambridge University, UK.

Spiegelhalter, D. J., Best, N. G., Carlin, B. P., & der Linde, A. V. (2002). Bayesian measures of model complexity and fit (with discussion). J. R. Statist. Soc.(64),583-639.

Weisberg, S. (2014). Applied linear regression(4th ed.). Hoboken NJ: Wiley.



How to Cite

BRAGANÇA, R. G. Bayesian inference applied to linear regression model and spatial model: An approach on the covariance structure between geostatistical data. Research, Society and Development, [S. l.], v. 10, n. 1, p. e31910111890, 2021. DOI: 10.33448/rsd-v10i1.11890. Disponível em: Acesso em: 28 feb. 2021.



Exact and Earth Sciences