An extension of geographically weighted regression with flexible bandwidths
An extension of geographically weighted regression with flexible bandwidths
复制标题
DOI：
--
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
Wenbai Yang
中科院分区：
文献类型：
--
作者：
Wenbai Yang
地理加权回归（GWR）（Brunsdon et al. 1996; Fotheringham et al. 2002）是一种用于模拟变量之间局部空间关系的有用技术。广义加权回归的基本思想是，在模型校正点附近的观测值对回归系数的估计比在模型校正点以外的观测值有更大的影响。标准GWR模型采用单一带宽来控制该影响中的距离衰减。然而，在实践中，这样的均匀带宽可能不足以反映因变量和自变量之间的关系中的复杂空间变化。为了得到一个更真实的模型，本文对GWR进行了扩展，在GWR中发现了灵活的带宽，提供了在不同空间尺度上变化的系数表面。在模拟数据集上进行了实验，对模型进行了检验。
Geographically weighted regression (GWR) (Brunsdon et al. 1996; Fotheringham et al. 2002) is a useful technique for modelling local spatial relati onships between variables. The essential idea of GWR is that observations near to a model calibratio n p int have more influence in the estimation of regression coefficients than observations farther a way do. The standard GWR model employs a single bandwidth to control the distance-decay in this inf luence. In practice however, such a uniform bandwidth may not be sufficient in reflecting compl ex spatial variations in relationships between dependent and independent variables. In an attempt to roduce a more realistic model, this paper develops an extension to GWR, where flexible bandwi dths are found providing coefficient surfaces that vary at different spatial scales. Experiments are carried out on simulated datasets to test the m odel.