County-level City Real Estate Price Forecasting System Based On Association Rules

In recent years,real estate prices have risen rapidly,and the rapid development of real estate industry has become a new growth point of local economy.At the same time,three-and fourth-tier small cities,especially county-level cities,have high real estate inventories,which are a great danger to industries such as construction industry.Data mining is a process of extracting information and knowledge that people don't know,but potentially useful,from a large number of random data.The application of data mining technology to price forecast in real estate sector can help enterprises find valuable information and adapt real estate price positioning to market.The application of data mining technology to price forecast in real estate sector can help enterprises find valuable information and adapt real estate price positioning to market.Only by establishing a set of related database of real estate enterprises,the real estate price fluctuations,and use the data mining technology,in order to better solve the problem of low efficiency of real estate enterprise management.Ability to comprehensively improve the operation efficiency of enterprises.To realize the sustainable,stable and coordinated development of the real estate industry and the entire national economy.At present,the research of existing data mining in real estate price forecasting can be divided into: the prediction model of multiple linear regression algorithm,multiple linear regression method.It is aimed at the development of the real estate price and the floating relationship between the price and the price,and clearly defined the model reference value of multiple linear regression.To determine the future trend of real estate prices.Defect is very obvious,but the class model due to the complexity of influence,interaction between reason and interfere with each other,the price fluctuation has certain periodicity,uncertainty,and regional characteristics.Impact prediction accuracy.Because data mining has little research on real estate price forecast in small cities.Due to the obvious characteristics of the real estate area,this paper aims to realize the forecast of real estate prices in the property market through the data mining of the real estate history transaction price.Collect real estate historical price data as the source data,standardization processing source data,set up the smallest degree of confidence of association rules algorithm in data mining and the minimum support,to improve the apriori algorithm,prediction model of mining design.Get the association rules that affect price and influence price.According to the association rules,combining prediction case attributes of the featureselection of similar cases,the corresponding correlation degree between attributes,and then use averaging method of fuzzy mathematics to calculate the prediction case price.The prediction error was 3% after the instance test.Meet the forecast accuracy requirement.