Abstract
With the commercialization of housing and the deepening of urbanization in China, housing prices are having increasing influence on the land market, and thus indirectly affecting urban development. As various spatial features of an urban housing property directly affect its price, the study of this connection has significance for urban planning. The present study uses mainly open internet data of housing prices, supplemented by other data sources, to identify the spatial features of housing prices and the influence factors in a case study city, Wuhan. Methods employed in the study include the hedonic linear regression model, the geographically weighted regression (GWR) model and the artificial neural network (ANN) model, etc. Progress is made in the following two aspects: first, when calculating the influence factors, hierarchical values for accessibility variables of certain public facilities are used instead of simple Euclidean distance and the results shows a better model fit; second, the ANN model shows the best fit in the study, and while the three models all show respective strengths, the combined use of all models offers the possibility of a more comprehensive analysis of the influence factors of housing prices.
Highlights
With the commercialization of housing and the deepening of urbanization in China, housing prices show an increasing influence on the land market, and affect urban planning and urban development [1]
In the urban planning discipline which directly guides the spatial development of cities, attention should be given to the study of the spatial features and the mechanism of how they influence housing price
In studies based on the hedonic model, one assumption is that all influence factors have a constant influence regardless of their different geospatial locations, while in reality, factors such as neighborhood feature and accessibility usually have a strong autocorrelation
Summary
With the commercialization of housing and the deepening of urbanization in China, housing prices show an increasing influence on the land market, and affect urban planning and urban development [1]. In the hedonic regression model, housing price does not exhibit a simple linear relationship with its influence factors [12]. Some scholars proposed the geographically weighted regression (GWR) model [19,20] in which the parameters and their weights can be adjust locally in accordance with their spatial locations To date, this model has been widely applied in many fields such as geography [21], economics [22], environmental science [23] and epidemiology [24]. Various works of research on housing price show that in studies of the same influence factor, results using the GWR model show a better fit than the simple hedonic model [25,26]. Such data sources significantly increase the number of samples with improved accuracy and timeliness [35]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
