The Application Of Nonparametric Regression In Hedonic Price Appraisal Method Of Commodity Housing

Regardless of speaking of the government or the consumer, the real estate all is the important wealth, therefore it needs accurately to appraise the value of real estate, for the use of the property tax, real estate business and mortgage loan and so on. But the real estate is a kind of heterogeneity commodity, and each real estate has obvious difference between the property characteristics, therefore how each real estate does carry on the science and reasonable appraisal needs objectively and fine method.Since the 1980's, the oversea scholars started to use the Hedonic Price Model to carry on the appraisal to real estate price, the Hedonic Price Model has become the main tool of real estate appraisal between the European and American countries now. The Hedonic Price Model is a kind of little subjective color method, it carries on price appraisal various from the concealed price embarks which real estate characteristics has, so it avoids appraisal deviation which the personnel subjective reason caused. But the Hedonic Price Model can meet the questions of the function hypothesis, the coefficient mark which not anticipated symbol and the spatial autocorrelation, in recent years the overseas scholar have started the non-parameter regression application in the real estate appraisal, and obtained the ideal appraisal effect.Based on summarizing the domestic and foreign correlations research empirical, this article takes the Hangzhou commodity housing as the example, establishes the Hedonic Price Model of commodity housing, and connects 2430 data samples to the Hedonic Price Model which first uses the traditional OLS method in four kind of functions form to carry on the regression analysis and most major function choice, then extracts 300 samples to conduct the commodity housing price appraisal research with parameter regression model, and extracts 100 samples to carry on the appraisal forecast effect to examine, finally uses nonparameter regression model to analysis further regression, and compares the result which obtained from parameter regression and nonparameter regression , thus obtains a kind of more reasonable method in Hedonic Price Model application to carry on the commodity housing price appraisal.