431
Views
0
CrossRef citations to date
0
Altmetric
Articles

A Bayesian Implementation of the Multiscale Geographically Weighted Regression Model with INLA

ORCID Icon &
Pages 1501-1515 | Received 03 Oct 2022, Accepted 07 Jan 2023, Published online: 05 May 2023
 

Abstract

The multiscale geographically weighted regression (MGWR) model is an important extension of the classical geographically weighted regression (GWR) model that can be used to explore the spatial nonstationarity of the regression relationship in spatial analysis, but also allows for different scales on conditional relationships between response and different predictors. A Bayesian version of the MGWR model is proposed to obtain estimates of the spatially varying coefficients and the bandwidths simultaneously. The hierarchical form of the Bayesian MGWR model has attractive features, including obtaining posterior estimates of the bandwidths and local parameters simultaneously, and their uncertainty can be easily measured. For Bayesian posterior inference, an efficient algorithm based on integrated nested Laplace approximation is introduced to provide a great alternative of the classical Markov chain Monte Carlo algorithm under the Bayesian framework. The performance of the proposed method is evaluated through simulation study, and it is shown that the proposed approach can correctly identify the differences between scales of parameter surfaces and also obtain precise posterior estimates. Finally, for illustration, this approach is used to analyze monthly housing cost data in the state of Georgia.

多尺度地理加权回归(MGWR)模型是传统地理加权回归(GWR)模型的重要扩展, 它可以在空间分析中探索回归关系的空间非平稳性, 可以考虑响应和预测因子之间的多尺度条件关系。本文提出了贝叶斯版本的MGWR模型, 旨在同时估计空间变化的系数和带宽。贝叶斯MGWR模型的分层形式具有以下优势:能同时获得带宽和局部参数的后验估计、便于评估不确定性。贝叶斯后验推理引入了集成嵌套拉普拉斯近似的高效算法, 为贝叶斯框架的传统马尔可夫链蒙特卡罗算法提供了优良的替代方案。模拟研究评估了该方法的性能。结果表明, 该方法能正确识别不同尺度的参数曲面差异, 并获得精确的后验估计。作为例证, 该方法被用于分析美国佐治亚州的月住房成本数据。

El modelo multiescalar de regresión geográficamente ponderada (MGWR) es una extensión importante del clásico modelo de regresión geográficamente ponderada (GWR) que puede usarse para explorar la no estacionalidad espacial de la relación de regresión en el análisis espacial, pero que además permite diferentes escalas en las relaciones condicionales entre la respuesta y los diferentes predictores. Se propone una versión bayesiana del modelo MGWR para obtener estimativos de los coeficientes que varían espacialmente y de los anchos de banda de manera simultánea. La forma jerárquica del modelo bayesiano MGWR tiene rasgos atractivos, incluida la obtención simultánea de estimativos posteriores de los anchos de banda y de los parámetros locales, y su incertidumbre puede medirse con facilidad. Para la inferencia bayesiana posterior se introduce un algoritmo eficiente basado en la aproximación integrada de Laplace anidada para proveer una importante alternativa del algoritmo Markov de la clásica cadena de Monte Carlo, bajo el marco bayesiano. Se evalúa el desempeño del método propuesto por medio de un estudio de simulación, y se muestra que el enfoque sugerido puede identificar correctamente las diferencias entre las escalas de las superficies de los parámetros y obtener también estimativos posteriores con precisión. Finalmente, a manera de ilustración, este enfoque se usa para analizar datos mensuales del costo de vivienda en el estado de Georgia.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

The authors’ research was supported by Ministry of Education in China Youth Foundation Project of Humanities and Social Sciences (21YJC910006), Shenzhen Science and Technology Program (20200812101943002), and Guangdong Basic and Applied Basic Research Foundation (2021A1515110220, 2023A1515011367).

Notes on contributors

Zhihua Ma

ZHIHUA MA is an Assistant Professor in the School of Economics, Shenzhen University, Guangdong, China. E-mail: [email protected]. Her research interests include Bayesian approaches for spatial data analysis and missing data problems.

Zhelin Huang

ZHELIN HUANG [corresponding] is an Assistant Professor in the School of Economics, Shenzhen University, Guangdong, China. E-mail: [email protected]. His research interests include time series analysis and machine learning methods.

Log in via your institution

Log in to Taylor & Francis Online

PDF download + Online access

  • 48 hours access to article PDF & online version
  • Article PDF can be downloaded
  • Article PDF can be printed
USD 53.00 Add to cart

Issue Purchase

  • 30 days online access to complete issue
  • Article PDFs can be downloaded
  • Article PDFs can be printed
USD 312.00 Add to cart

* Local tax will be added as applicable

Related Research

People also read lists articles that other readers of this article have read.

Recommended articles lists articles that we recommend and is powered by our AI driven recommendation engine.

Cited by lists all citing articles based on Crossref citations.
Articles with the Crossref icon will open in a new tab.