Early Warning System Research Based On Neural Networks For Nanjing Real Estate Market

Real estate is an important part of national economy, which has cycling characteristics. Too large fluctuation contradicts with the continuing healthy and stable development of the real estate economy. At the same time, as a subsystem of the society economy system, real estate has non-linear complex traits.The purpose of this article is: in order to prevent the real estate economy from the nonnormal fluctuation and keep it developing healthy and stable, through establishing an advanced and scientific early warning system based on nonlinear real estate system.First of all, this paper summarizes the real estate early warning system in nanjing, gives the empirical research outcome of the two finished early warning models and methods, selects the indexs of early warning based on neural networks model for real estate early warning and analyses the three models and methods compared.Secondly, based on the neural networks and real estate earning warning theory, this paper utilizes the mature fuction of forecast and model identification of the neural networks, then focuses on the model and mothod of real estate forecasting&early warning based on the neural networks theory. With the real estate market in nanjing taken as a subject, the above mentioned model and method is employed to do empirical research..Finally, based on the results of the analysis of the three early warning models, this paper carries out the integrated early warning result and gives suggestions for policy regulation. The result of the analysis of early warning is in accordance with the practical situation of the real estate development in Nanjing, this indicates that the real estate early warning system established in the research is feasible, with full theory analysis and good practical value, provides scientific foundation for guiding and controlling the real estate markets.