多元回归和神经网络在武汉市房地产预测中的应用比较-应用统计专业论文.docxVIP

华中科技大学硕士学位论文 华 中 科 技 大 学 硕 士 学 位 论 文 II II ABSTRACT The real estate is a complex integrated systems engineering, which is therelationship between the lifeblood of the national economy. The real estate market has cycling characteristics. Too large fluctuation will make aggainst the continuing healthy and stable development of the real estate market, thereby affecting the sustainable development of the national economy. Prediction of the real estate market forecast is the volatility of real estate-related indicators. the establishment of the advanced and scientific real estate forecast system can effectively avoid the real estate market from the non-normal fluctuation and keep it developing healthy and stable. This is the purpose of this study. This article first introduces the prediction model, multiple regression model and neural network model. It includes principles, algorithms and testing of these two models, as well as BP neural network model which applied in this paper. Secondly, This article utilizes the mature use of multiple regression model and BP model in predicting and pattern recognition. With the real estate market in Wuhan as a object, SPSS, Matlab statistical data mining tools are used according to Wuhan City real estate market relevant data to establish the corresponding model. This article focuses on the model and the method of BP neural network theory of real estate based on Prediction and discusses two models of the real estate forecast. Finally, based on the result of the analysis and the comparison of two different prediction methods, this paper carries out the consolidated forecast conclusions and proposes relevant policies and regulation. The result of analysis of forecast is in accordance whith the pratical situation of the real estate development in Wuhan. This indicates that the model established in the research is feasible, with full theory analysis andgood practical value. Through the analysis on the prediction results we can found that th