房地产特征价格模型应用的比较研究-数量经济学专业论文.docxVIP

Abstract Real estate is a kind of commodity with heterogeneity, The values of properties are different in the real estate market, resulting in the difficulties in accurate assessment of the real estate. The existing real estate valuation method used in the process of real estate appraisal inevitably have some subjective factors, the valuation of the real estate have great dependence with the subjective judgement of the assessment staff. Foreign scholars commonly use the Hedonic model to assess real estate. Hedonic model can fully consider all the factors affecting the price of real estate, including more market information and the prices can be formed on the objective evaluation and avoid the subjective judgement in real estate assessment The most commonly used aproach to establish the Hedonic model is parameter regression,which can explain the impacts of characteristics of real estate to the price by estimating the parameters of the model.However, the parametric model in the practical application is often specified error if the assumptions can not be met, leading to the results unreliable.On the basis of the domestic and abroad related studies, introduct the nonparametric regression and semiparametric regression in Hedonic model .Use the second-hand market data of ordinary residential of Haikou City to establish Hedonic model with nonparametric regression and semiparametric regression. In nonparametric regression, use the tools of kernel, k -neighbors and local linear regression to obtain a better estimate effect. Use the four-stage method in the semiparametric regression model . By comparing the results of the regressions,try to get a better fitted and more reasonable approach. Key words: real estate Hedonic model parametric regression nonparametric regression semiparametric regression 目 录 摘 要............................................................................................................... I HYPERLINK \l _bookmark0 Abstract ........................