As the saying goes, nothing can be more important than the location in the characteristics of real estate, it fully illustrates the importance of the location to real estate prices. Due to the immobilization of residence space and the unbalanced regional development of cities, large variations in price exist in different residence. To study the diversity of price in different mansions can provide decision-making basis of policy formulation for the government, the scheme selection for the real-estate developer and the house purchase selection for the citizen.This paper takes Hefei as the research area,4954 samples of price are collected and the variations and influence factor of the price in Hefei city by Geographical Weighted Regression (GWR) is studied. The price movement of Hefei is analyzed by adopting exploratory spatial data analysis. It turned out that the price move parabolic ally between north to south or east to west. Then, the spatial autocorrelation analysis is used, from which the clustering of the price is shown, and characteristics is described by interpolation technology. Ten influence factors are chosen and the GWR is established, and then the mechanism of the influence factor is investigated deeply, moreover, the hedonic model is compared with the GWR. The comparison result show the advantage of GWR, and the result indicate that the influences of price work in both directions. Meanwhile, the visualization model is build by interpolation which can visualize the effect of the price clearly. |