Estimation Of Spatial Autoregressive Models Research Articles

Spatial ARMA models and its applications to image filtering

The objective of this review paper is to summarize the main properties of the spatial ARMA models and describe some of the well-known methods used in image filtering based on estimation of spatial autoregressive models. A new proposal based on robust RA estimation is also presented. Previous studies have shown that under additive outliers the RA estimator is resistant to a small percentage of contamination and behaves better than the LS, M, and GM estimators. A discussion about how well these models fit to a digital image is presented. Some applications using real images are presented to illustrate how an image is filtered in practice.

Profile quasi-maximum likelihood estimation of partially linear spatial autoregressive models

We propose profile quasi-maximum likelihood estimation of spatial autoregressive models that are partially linear. The rate of convergence of the spatial parameter estimator depends on some general features of the spatial weight matrix of the model. The estimators of other finite-dimensional parameters in the model have the regular n -rate of convergence and the estimator of the nonparametric component is consistent but with different restrictions on the choice of bandwidth parameter associated with different natures of the spatial weights. Monte Carlo simulations verify our theory and indicate that our estimators perform reasonably well in finite samples.

GMM estimation of spatial autoregressive models with unknown heteroskedasticity

In the presence of heteroskedastic disturbances, the MLE for the SAR models without taking into account the heteroskedasticity is generally inconsistent. The 2SLS estimates can have large variances and biases for cases where regressors do not have strong effects. In contrast, GMM estimators obtained from certain moment conditions can be robust. Asymptotically valid inferences can be drawn with consistently estimated covariance matrices. Efficiency can be improved by constructing the optimal weighted estimation. The approaches are applied to the study of county teenage pregnancy rates. The empirical results show a strong spatial convergence among county teenage pregnancy rates.

Specification and estimation of spatial autoregressive models with autoregressive and heteroskedastic disturbances

This study develops a methodology of inference for a widely used Cliff–Ord type spatial model containing spatial lags in the dependent variable, exogenous variables, and the disturbance terms, while allowing for unknown heteroskedasticity in the innovations. We first generalize the GMM estimator suggested in Kelejian and Prucha (1998, 1999) for the spatial autoregressive parameter in the disturbance process. We also define IV estimators for the regression parameters of the model and give results concerning the joint asymptotic distribution of those estimators and the GMM estimator. Much of the theory is kept general to cover a wide range of settings.

Spatial interaction of crime incidents in Japan

We analyze the development of 18 types of criminal records in Japan for the period 1991–2001 across 47 prefectures with spatial lag and spatio-temporal heteroscedasticity. We explore the hypothesis that crime data are related to socio-economic variables in Japan. We extend the Bayesian approach of LeSage [J.P. LeSage, Bayesian estimation of spatial autoregressive models, Int. Regional Sci. Rev. 20 (1997) 113–129] for spatio-temporal Bayesian models. Additionally we analyze unobserved heteroscedasticity in the panel model by variance inflation factors as in Geweke [J. Geweke, Bayesian treatment of the independent Student- t linear model, J. Appl. Econ. 8 (1993) 19–40]. Positive and significant spatial dependencies can be found for 12 types of crimes and the influence of the socio-economic variables varies over the type of crimes.

Specification and Estimation of Spatial Autoregressive Models with Autoregressive and Heteroskedastic Disturbances

One important goal of this study is to develop a methodology of inference for a widely used Cliff-Ord type spatial model containing spatial lags in the dependent variable, exogenous variables, and the disturbance terms, while allowing for unknown heteroskedasticity in the innovations. We first generalize the generalized moments (GM) estimator suggested in Kelejian and Prucha (1998,1999) for the spatial autoregressive parameter in the disturbance process. We prove the consistency of our estimator; unlike in our earlier paper we also determine its asymptotic distribution, and discuss issues of efficiency. We then define instrumental variable (IV) estimators for the regression parameters of the model and give results concerning the joint asymptotic distribution of those estimators and the GM estimator under reasonable conditions. Much of the theory is kept general to cover a wide range of settings. We note the estimation theory developed by Kelejian and Prucha (1998, 1999) for GM and IV estimators and by Lee (2004) for the quasi-maximum likelihood estimator under the assumption of homoskedastic innovations does not carry over to the case of heteroskedastic innovations. The paper also provides a critical discussion of the usual specification of the parameter space.

Chebyshev approximation of log-determinants of spatial weight matrices

To cope with the increased sample sizes stemming from geocoding and other technological innovations, this paper introduces an O( n) approximation to the log-determinant term required for likelihood-based estimation of spatial autoregressive models. It takes as a point of departure Martin's (1993) Taylor series approximation based on traces of powers of the spatial weight matrix. Using a Chebyshev approximation along with techniques to efficiently compute the initial matrix power traces results in an extremely fast approximation along with bounds on the true value of the log-determinant. Using this approach, it takes less than a second to compute the approximate log-determinant of an 890,091×890,091 matrix. This represents a tremendous increase in speed relative to exact computation that should allow researchers to explore much larger problems and facilitate spatial specification searches.