Local Linear Geographically Weighted Regression Analysis on the Urban Housing Price: A Case Study of Huangshi City, Hubei Province

This paper aims to analyze the spatial variations of urban housing price and its impact factors using local linear geographically weighted regression(LLGWR) model. For purpose of real estate market management and land use policy making scientifically, Huangshi City, Hubei Province, a typical land-scarce, resource-based and multi-group city, is selected as the study area. Additionally, 193,00 housing units and 398 buildings are collected as modeling samples. Based on the general and spatial statistics, the floor numbers, the plot ratio, the greening ratio, the level of property management, the distance to nearest urban center, and the year of sale are selected as explanatory variables, constructing the model and carries on the analysis and interpretation. The modeling results indicate that, compared with ordinary linear regression(OLS), GWR and LLGWR are more suitable for interpreting urban housing price, and LLGWR is better than GWR. The housing price is substantially affected by the year of sale, plot ratio and the urban geographic location; and it is also related to the building height and greening ratio. However, they have very different contributions in different urban functional zones. There are three findings from this research: 1) LLGWR model, using unbiased estimation of coefficient function and error variance, can improve estimation and prediction accuracy of the urban housing price; 2) Macro market trend is the key factors affecting the housing price, but in different geographical location, housing price growth trend is obvious difference; 3) The spatial distribution of real estate development is closely related to land use planning and land supply policy, however, the relationship between land price and housing price is not obvious in the study area.