Research On Housing Price Data Of Chengdu City Based On Kalman Filter

In our country,real estate is an important support for economic growth,and the real estate industry is very important.According to the wealth effect,in a typical country,the market value of real estate is generally two to three times the country’s annual GDP.As of the latest news,the total market value of my country’s real estate has reached 65 trillion U.S.dollars after conversion,while the United States reached 33.6 trillion U.S.dollars and Japan reached 10.8 trillion U.S.dollars during the same period.My country’s GDP is around 100 trillion yuan,after conversion,the ratio of total housing value to GDP is between 4-5.Before and after the 2008 financial crisis,the ratio of US housing market value to GDP was as high as 169%.In 1990,Japan’s real estate Before the bubble,the ratio of real estate market value to GDP reached a maximum of 391%.From the perspective of several real estate bubble crises in history,the impact of the real estate crash has been large and far-reaching,including financial crises,economic crises,social crises,and political crises.The impact in the world has made the global economy still not out of the shadows so far.At the stage when my country’s economy changes from high growth to high quality,the growth of the total real estate value also needs to be strictly controlled by the state.Domestic and foreign scholars’ research on housing prices is divided into several categories,including prediction of housing prices,research on housing price bubbles,and discussions on reasonable housing prices,explaining changes in housing prices from two aspects: statistical models and economic significance.However,the data sources of traditional research rely on offline and are highly subjective.With the advent of the era of big data,online powerful databases using existing data sources are more accurate,efficient and granular.Based on the huge amount of online data,the data source selected in this paper is the full amount of housing information data in Chengdu crawled from the housing transaction platform.The location of the house is divided into different areas under different areas to ensure that the particle size is fine enough to conduct research and analysis at different spatial scales.In terms of model method,this paper uses the state space Kalman filter method and the method of multi-sensor information fusion to calculate the potential house prices that change with time in different areas in different areas of Chengdu,and obtain the difference between the measured value and the listed house price and discuss the differences.Then,the house prices of different areas in the same area are fused to obtain a relatively stable potential house price without noise disturbance.The state space Kalman filter algorithm has been fully applied in econometrics,but the algorithm of multi-sensor information fusion is still seldom used in the field of financial economy,which is also a major innovation of this paper.Through the application of state-space Kalman filtering and multi-sensor information fusion,the first conclusion is that the potential house price obtained after removing noise disturbance is very stable,and the average value can be taken as the guide reference price for the region;Some of the potential housing prices in districts and counties are higher than the listed price,and some are lower than the listed price.The lower potential housing price mainly occurs in the inner circle,whereas the higher situation mainly occurs in the outer circle.Therefore,it is necessary to combine the conditions of each region.according to the practical problems of various regions,and in response to national policies,both “stop rising”and “steady falling” of housing prices should be balanced.