Price Forecasting Model Study Of Urban Real Estate Based On Multi Factor LOGISTIC

The price of real estate has been the hottest issues social concerns. As Pillar Industry, the solving of the housing problems is a direct impact of the national economy's development, and unity of the society. And the core of the housing problems is housing price. In recent years, the economic of our country has developed quickly. The surplus of the fluidity is significantly. The highly enthusiasm of the housing buyers promote the housing price rising continuously. While since September 2007, there are many new situations. The sales began to decline and the housing price of some cities began to fall. The trend of the housing price caused concern of the whole country. However, the research about housing price has just started. This paper tried to make some exploration from this fact.This paper starts from the aspects that impact the price of real estate. Had a system investigation of domestic and international research on housing price forecasting. At the same time, it had a comprehensive understanding about the performance of housing price. It oriented the research of this paper. The paper sums up big4 theories about the housing price and analysis the supply and demand theory of real estate. It had a theoretical derivation of the housing demand function and the housing supply function. Then, it had a history retrospectively to Chinese housing price and analysis the housing supply and demand status of our country. Finally, it is based on panel data of 35 cities in 2002-2006, and used traditional multi linear regression model (LRM) as contrast to carry on price forecasting model investigation of real estate with multi-factor LOGISTIC method. The results show that, the veracity of model No. 1 based on least square multiple regression analysis reaches 81.6%; the veracity of model No. 2 based on LOGISTIC regression analysis reaches 86.2%. Therefore, it is believed that the model based on LOGISTIC regression analysis has greater advantage than that of traditional multi linear regression model (LRM) at the price forecasting of real estate. At the same time, the research indicates that the following aspects for the year : residential housing sale price index, growth rate of sale area and floor space completed, interest rate changing percentage are main and important factors to impact the residential housing price for the next year.