The Prediction Of House Price Based On Multiple Regression Analysis And Support Vector Machine

After nearly forty years of development,China's real estate industry has reached to 18% of China's total GDP.The whole industry has been closely linked with the national economy.In the forty year history of industry development,China's real estate industry has its own laws of development.Moreover,the development of the whole industry is gradually divorced from the original single real estate attributes,and shows a trend of real estate finance.Against this background,the combination of some cities and new formats has led to the development of cities in a more livable environment.The successful experience of foreign countries in property tax collection system is worth summarizing and drawn from China.This thesis takes the prediction of house price as the main object of study.First,studying the main factors of influencing house price,we establish a set of linear regression and non-linear regression mathematical models,so as to provide a basis for explaining house price fluctuation.Second,the support vector machine method is used as a comparison of mathematical statistics methods,and a better prediction result is reached.Finally,it introduces several typical design structures of real estate financial products,as well as several types of financial risks.