An Study Of Urban Fringe Residential Land Price Evaluation Based On The Kernel Discriminant Analysis And Support Vector Machine

As China's urbanization process deepened, China's urbanization work has been an engine to solve the employment and expand domestic demand. It is the key way to speed up transferring rural population and eliminating the urban-rural dual structure, and its position in the national economy is rising up. In this process, the land is a scarce resource and its market value is reflected immediately. Land price has become an important reference factor for government to formulate the land policy. Urban fringe is the focal zone of urbanization, and housing is the most important means of subsistence and the most common investment. Residential land price in urban fringe will inevitably affect the living and employment of the rural transfer population, the industrial development in urban fringe and the investment financing motives of financial capital and financial capital. So, the land price in urban fringe which can reflect the supply and demand of residential land has become the focus of the scholars. But because of the specificity of its own, the affect factors of the residential land price in urban fringe are different from those of urban or rural. It complexity brings a challenge for the evaluation of the residential land price in urban fringe. Well, what kind of the index system affecting the residential land price in urban fringe, how to recognize the residential land value correctly and how to evaluate the residential land price in urban fringe scientifically and accurately has become an important problem which should be solved now in the process of urbanization. So, making a research to evaluate the residential land price in urban fringe is important in theory to further improve our land policy system and has practical significance in practice to promote the economic development of urban and rural, and to establish a integration price system of land in urban and rural.First, we form an index system that contains the affect factors of residential land price in urban fringe by summarizing the urban fringe characteristics and using the derived demand theory. It provides a basis for the quantitative analysis, that's used in the variable factors of residential land price in urban fringe next. Second, the feature extraction based on the kernel discriminant analysis is used to extract the main affect factors from the index system formed before. And then, the support vector machine is used to establish the regression model for the residential land price evaluation in urban fringe. At last, we select the land transaction cases in Chongqing, take Chongqing urban fringe for example, and make an empirical study of its residential land price. The research result shows that this model can be used to obtain the effective affect factors of the residential land price in urban fringe and make an objective assessment of it.