Spatial Regression and the Bayesian Filter

TL;DR

This paper introduces Bayesian spatial filtering as a practical approach for spatial regression, balancing the simplicity of non-spatial models with the complexity of traditional spatial models, supported by simulation results.

Contribution

It proposes Bayesian spatial filtering as a new method that offers a middle ground between existing spatial regression models, with demonstrated desirable properties.

Findings

Bayesian spatial filtering performs well in simulations.

The method offers a computationally efficient alternative.

It provides a flexible framework for spatial regression analysis.

Abstract

Regression for spatially dependent outcomes poses many challenges, for inference and for computation. Non-spatial models and traditional spatial mixed-effects models each have their advantages and disadvantages, making it difficult for practitioners to determine how to carry out a spatial regression analysis. We discuss the data-generating mechanisms implicitly assumed by various popular spatial regression models, and discuss the implications of these assumptions. We propose Bayesian spatial filtering as an approximate middle way between non-spatial models and traditional spatial mixed models. We show by simulation that our Bayesian spatial filtering model has several desirable properties and hence may be a useful addition to a spatial statistician's toolkit.