292
Views
0
CrossRef citations to date
0
Altmetric
Articles

l0-Norm Variable Adaptive Selection for Geographically Weighted Regression Model

ORCID Icon, ORCID Icon &
Pages 1190-1206 | Received 22 Jul 2022, Accepted 21 Nov 2022, Published online: 02 Mar 2023
 

Abstract

A geographically weighted regression (GWR) model with fewer explanatory variables and higher prediction accuracy is required in spatial analysis and other practical applications. This article proposes an l0-norm variable adaptive selection method to enhance performances of a GWR by simultaneously performing model selection and coefficient optimization. Specifically, we formulate a regularized GWR model with an additional l0-norm constraint to shrink those unimportant regression coefficients toward zero and propose an adaptive variable selection algorithm by iteratively distinguishing the important variables from the variable set. At each location, the best variable subset and optimizing coefficient estimations are simultaneously achieved under the l0-GWR framework. Moreover, two novel criteria, the modified Bayesian information criterion and the interpretability of coefficient symbol, which specify the variable selection and model interpretation, respectively, are also introduced to improve the performance of the l0-GWR. Experiments on both simulated and actual data sets demonstrate that the proposed algorithm can significantly improve the estimation accuracy of coefficients and can also enhance the interpretative ability of the established model.

空间分析等应用要求地理加权回归(GWR)模型具有更少的解释变量和更高的预测精度。本文提出l0范数变量自适应选择方法, 同时进行模型选择和参数优化以提高GWR的性能。我们建立具有l0范数约束的正则化GWR模型, 将次要回归参数缩小到零;提出自适应变量选择算法, 通过迭代方法从变量集合中筛选重要变量。基于l0-GWR框架, 在每个空间位置上同时实现最佳变量子集和参数优化估计。为了提高l0-GWR的性能, 还引入新的变量选择准则和模型解释准则(即, 修订的贝叶斯信息准则和参数可释性准则)。基于模拟数据和实际数据的实验表明, 本算法可以显著提高参数估计的精度、增强模型的解释能力。

En análisis espacial y en otras aplicaciones prácticas se requiere un modelo de regresión geográficamente ponderada (GWR) con menos variables explicativas y una mayor exactitud en la predicción. En este artículo se propone un método de selección adaptativa de variables 10-norm, para fortalecer los desempeños de una GWR, realizando simultáneamente la selección del modelo y la optimización de los coeficientes. Específicamente, formulamos un modelo GWR regularizado, con una constricción 10-norm adicional para contraer aquellos coeficientes de regresión sin importancia hacia cero, y proponemos un algoritmo adaptativo de selección de variables, distinguiendo iterativamente las variables importantes del conjunto de variables. En cada una de las ubicaciones, se consiguen simultáneamente el mejor subconjunto de variables y la optimización de los estimativos de coeficientes bajo el marco. Por lo demás, también se introdujeron dos criterios novedosos, el de información bayesiana modificado y la interpretabilidad del símbolo del coeficiente, respectivamente, para mejorar el desempeño del 10-GWR. Los experimentos efectuados con conjuntos de datos simulados y con conjuntos de datos reales demuestran que el algoritmo propuesto puede mejorar de modo significativo la exactitud del estimativo de los coeficientes y puede también fortalecer la capacidad interpretativa del modelo establecido.

Acknowledgments

We would like to take this chance to thank the anonymous reviewers for their insightful comments that have been very helpful in improving this article.

Additional information

Funding

Funding was provided by the Natural Science Foundation of China (Grant No. 41961055, U1811464) and Scientific Research Fund of Hunan Provincial Education Department [Grant No. 22A0498].

Notes on contributors

Bo Wu

BO WU is a Professor in the Department of Geography and Environment at Jiangxi Normal University, Nanchang, Jiangxi Province, 330022, China. E-mail: [email protected]. His research interests include spatiotemporal data analysis, GeoAI, and remote sensing image processing.

Jinbiao Yan

JINBIAO YAN is a PhD Candidate in the Department of Geography and Environment at Jiangxi Normal University, Nanchang, Jiangxi Province, 330022, China. E-mail: [email protected]. His research interest is spatial statistical analysis.

Kai Cao

KAI CAO is a Professor in the School of Geographic Sciences, East China Normal University, Shanghai, 200241, P.R. China. E-mail:[email protected]. His research interests include spatial simulation and optimization, urban studies, and spatially integrated social science.

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.