Based On Gray Theory And Multiple Linear Regression Analysis Of Real Estate Forecasting Model And Its Empirical Analysis

As a commonly used multivariate statistical methods and for its principles is clear, model is simple and easy to use, classical multiple linear regression model has been a very wide range of applications in the industrial and agricultural production and scientific research. For example, in medical and health, weather forecasting, geological prospecting, and many other areas have to use the classical multiple linear regression model.Since the 1998 housing reform, China's real estate industry has been a rapidly development, and make a significant contribution to the national economy and gradually become the pillar industry of China's national economy. However, for its late start and development of imperfect, China's real estate market shows a more obviously up and downs of volatility and some other more typical features of a primary market. In this case, the Government would normally take some measures to guide the real estate to make the right investment.A city's average price of commercial housing sales is often affected by this city's residential land premium, average wages, and per capita disposable income of urban residents. Before making the decision of investing, realtors tend to predict the city's average price of commercial housing sales. A traditional prediction method is the classical multiple linear regression analysis, that is make the city's residential land premium, average wages, per capita disposable income of urban residents as independent variables, the average price of commodity housing sales as a responsible variable to carry out least squares estimation to predict the prices. That approach, while easy to understand, but there are some problems:First, it can not track the responsible variable's changes in real-time; Second, the classical multiple linear regression analysis model is sensitive for a few ill-sensitive data, often a result of a small amount of pathological data affect the fitting results; Third, classical multiple linear regression model needs large amounts of data. These issues are likely to lead to real estate to make a wrong decision analysis and affect the healthy development of China's national economy.To address the above problems, this paper presents a new real estate forecast model, which is based on gray theory and multiple linear regression prediction models to predict the right price. This method is first to go through the GM(1,1) model in the gray theory to predict the premium residential land, average wages, per capita disposable income of urban residents, and then use the classical multiple linear regression model to predict the housing prices. The advantages of this method is it can be avoided with a small amount of it not only ill-effect of data fitting, but also needs small amount of data.In the end of the paper, we choose Xi'an, Xianyang City's real estate as an analysis object and verified the real estate forecasting's accuracy, the results show that this method can be used to help the realtors forecasting commodity housing Sales average price in order to make the right decision analysis.