| This study examined the impact of public service provision on the spatiotemporal patterns of housing prices in Guangzhou City,Guangdong Province,and through modeling,use POI data to predict the housing prices of grid units without data.The research shows that the house price of Guangzhou shows the overall characteristics of high center and low surrounding,high South and low north.It decreases from the core urban areas(Yuexiu,Liwan,Haizhu,Tianhe)along the main traffic line to the surrounding urban areas,and the overall structure is single center circle plus radial space.The house prices in Guangzhou have obvious clustering characteristics,and there are significant differences in the clustering distribution between different urban areas.In different circles and radiation directions,public services that have significant impact on house prices are not the same.The impact of public service factors on house prices is affected by the spatial function differences of different regions.Based on machine learning,this paper studies the algorithm used to predict the house price of the grid unit with missing data.The absolute median error of the predicted house price is only 6291 yuan per square meter in the grid unit of every square kilometer.Considering that there are many factors involved in the grid unit(such as the house type,greening,etc.),the prediction effect is ideal.In general,the distribution of urban public service facilities in Guangzhou generally presents a pattern from the center to the suburbs,the difference of regional supply scale is large and inconsistent with the population distribution,while the differences in public service provision and spatial functions will further deepen the differentiation pattern of housing prices.Therefore,in the adjustment and control of housing prices in Guangzhou,we argue that a better understanding of the influence of public services on housing prices will help policymakers to optimize spatial functions and public services. |
PDF Full Text Request