于GWR模型的广州市住宅价格影响因素研究

以广州市“都会区”为例，基于2016~2019年住宅小区均价及POI数据，综合运用kriging插值、GWR模型对广州市住宅价格空间分布结构及其影响因素进行研究。研究结果表明:2016~2019年广州市住宅价格整体呈上升趋势，沿珠江水系发展的双核、多次中心、组团式的空间结构已基本成型；高速/快速路口、三甲医院、省一级小学对住宅价格的回归系数在中心城区与近郊区域间呈明显的正负差异，绿化率、地铁、房龄等单向影响因素回归系数强弱渐变；房龄、CBD、省一级小学、三甲医院、绿化率、休闲配套对中心城区住宅价格影响较大，而近郊区域主要受地铁、高速/快速路口的影响；较全局OLS模型，GWR模型的拟合优度大幅提高19.9%，能够更加精确地研究住宅价格影响因素的空间异质性。

Study on the Influencing Factors of Housing Price in Guangzhou Based on GWR Model

Taking Guangzhou metropolitan area as an example，based on the average price and POI data of residential quarters from 2016 to 2019，this paper comprehensively uses Kriging interpolation and GWR Model to study the spatial distribution structure of residential prices and its influencing factors in Guangzhou. The results show that:From 2016 to 2019，the housing price in Guangzhou is on the rise，and the spatial structure of dual-core，multi-center and cluster along the Pearl River system has been basically formed；The regression coefficients of expressway/fast intersection，third-class hospital and provincial primary school on the housing price show obvious positive and negative differences between the central urban area and the suburban area，while the regression coefficients of one-way influencing factors such as greening rate，subway and housing age gradually change；Housing age，CBD，provincial primary schools，third-class hospitals，green rate and leisure facilities have a greater impact on housing prices in central urban areas，while the suburban areas are mainly affected by subways and expressways/fast intersections；Compared with the global OLS model，the goodness of fit of the GWR model is significantly improved by 19.9%，which can more accurately study the spatial heterogeneity of influencing factors of housing price.