Modeling For Real Estate Forecasting And Early Warning System And Its Empirical Research In Qingdao

Real estate industry, which is playing a leading role in the development of national economy, shows more and more obviously the characteristic of cycle fluctuation. Its excessive fluctuation contradicts with the sustainable, healthy, and stable development of the real estate economy. Currently, there is still much work to do to perfect the real estate market operation mechanism of China, to form rational, orderly, competitive and efficient market operation system and to solve the current problems of the unsmooth transferring of information, the distortion of information data, the backward displaying means of market quotation and the blocking of the market transaction network. It has become a hot point in the academic cycles and arouses intensive concerns of the related policy-making apartments to do research on real estate early warning system, to set up the real estate early warning index system and systematically, scientifically and accurately define the secure region of real estate.To address this problem, the research of early warning models of real estate market is put forward in this dissertation which will make its theoretical contribution to the promotion of the healthy development of real estate industry. On the base of this research, empirical study is also done on the real estate market of Qingdao.Based on the research above, the main findings and the conclusions are mentioned as follows:First of all, based on the analysis of the current researches at home and abroad, this dissertation summarizes the basic concepts, principles and methods of forecasting and early warning of real estate economy. On the base of the analysis of the domestic real estate index systems, this dissertation selects and fixes on its indexes of early warning.Secondly, as a subsystem of the social and economic system, real estate shows complex non-linear characteristics. This dissertation aims at solving the nonlinear problem of the real estate system and establishing a more advanced and scientific early warning system in order to prevent the real estate market from the abnormal fluctuation and to maintain the sustainable, healthy and stable development of the real estate market. On the foundation of the present research, the dissertation systematically analyses the characteristics and functions of the real estate early warning, make an identification, prediction, diagnosis, monitoring and control of the key index in the process of real estate early warning and construct a theoretically and practically feasible real estate early warning system which provides the basis for the solution of the real estate early warning problems.Thirdly, based on neural networks theories and real estate earning warning theories, this dissertation introduces its forecasting and early warning system and develops its model of forecasting and early warning. By utilizing the neural networks which is maturely applied in the field of forecast and model recognition, the dissertation puts its emphases on the research of the models and methods of real estate forecasting and early warning. And based on these models this dissertation develops its own forecasting and early warning system of real estate market.Fourthly, based on the real estate early warning index system, this dissertation develops the LVQ-RBF neural networks model of forecasting and early warning. With high parallelism, global superiority, accuracy and applicability this model has overcome the deficiency of traditional early warning methods and highly improves the real estate early warning system's non-linearity, self studying ability, self adaptability and the ability to process large-scale concurrently distributed knowledge.Finally, based on the previous theoretical research findings, this dissertation empirically analyses of the real estate market in Qingdao and forms its comprehensive early warning theories and methods. Since the conclusion of the analysis of early warning in this dissertation is in accordance with the practical development of the real estate in Qingdao, the real estate early warning system established in this research is proved to be feasible, with full theory analysis and good practical value, and provide scientific foundation for guiding and controlling the real estate markets.