MRS Combination Model Research In House Price Index Forecasting

The thesis had analyzed the factors of the composition of house-price and built the house pricing model, therefore, it is possible to divide the composition of house price into two items: the policy economic command item and the cost index item, with these factors, we chose the reasonable replacement index according to it, whose dates can well to find and use for multiple regression model. And then, the thesis analyzed the relationship of these indexes with E-G method, and offered the antecedence steps to make multiple regression models with the related price index. Moreover, to enable the price forecasting model more reliable, the thesis had considered the characteristic property, and built the combination forecast model, solving with genetic algorithm, that the error is smaller.By analyzing the Price forecasting model, the thesis provided the main policy to regulate house market, and proved that it needs the marketing method to regulate the house price structure but not the man-made way. And the combination policy of house least and related tax is up to Mechanism Design.The innovation of the thesis is line in the following three points:1. The forecasting model's index come from the house pricing model; 2.Built the regression model on house-price index, which can forecast the change of the price; 3. Built the combination forecast model and solve by Genetic Algorithm; 4.Provided the house price regulation policy that meet the Mechanism Design.