The Modeling Analysis And Forecast Of Supply And Demand In The Real Estate Industry Based On Gray And Artificial Immune Algorithm

The real estate industry is one of the pillar industries of national economy, inrecent years, the rapid development of real estate industry made a great contributionto the national economy, it has a pivotal role in China’s economic and socialdevelopment. At the same time, the supply, demand and prices of real estate is closedrelated to people’s lives, and real estate has become one of the most importantlivelihood issues. Many scholars have studied the current supply, demand and pricesof real estate, but mostly from the perspective of policy and funding, the qualitativeanalysis is more than quantitative analysis. Through the establishment ofmathematical model to analysis the quantitative relation between the factors whichaffecting the development of real estate industry, accurately predicted the trend of thesupply, demand and prices of real estate, then provide decision support for effectivecontrol, it is a direction worth exploring.This thesis first study of the factors that affect the supply, demand and prices ofreal estate from the perspective of qualitative in-depth, then select the NationalBureau of Statistics of People’s Republic of China <In2013China StatisticalYearbook>, using multiple linear regression method carried on quantitative analysisto it. Finally, using the artificial immune method and grey prediction method and thenput forward an effective prediction model for the supply, demand and prices of realestate. The achievement of this research is mainly as follows:(1) On the basis of qualitative analysis and adopts the data which can beacquired, selected factors which influence on supply, demand and prices of real estatemost closely and used stepwise regression method for quantitative analysis. In viewof different sample data, using forward stepwise regression, backward stepwiseregression, Stepwise regression with a constant and stepwise regression withoutconstant for modeling, and compared the fitting accuracy and effectiveness of theobtained models.(2) Grey system theory focuses on the study of small sample, poor informationproblem, the real estate industry’s annual supply, demand and sales price series data is less, suitable for the use of grey prediction model. This thesis used the GM(1,1)model to forecast the supply, demand and prices of the real estate. In order to improvethe fitting precision of standard GM (1,1) model, immune clonal selection algorithmis introduced into the gray GM(1,1) model. This thesis proposed two optimizationalgorithms.(3) Standard GM(1,1) model using background which is taking the meancumulative number of adjacent, it does not accurately reflect changes in datasequences and lead to the predicted value and the actual value has a greater error. Thisthesis put forward an improved algorithm which use of immune clonal selectionalgorithm to optimize the parameter β.(4) Standard GM (1,1) model using the least square method to calculate theparameters a and b, but the least square method has many restrictions may lead to theaccuracy of predicted value is not high. This thesis use of immune clonal selectionalgorithm for optimization based on a, b to obtain more accurate a, b.Using the improved immune clone algorithm base GM (1,1) model modeling ofthe supply, demand and prices of real estate. The experimental results show that thetwo improved algorithms are better than the standard GM (1,1) algorithm for fittingprecision of the original data sequence. This thesis proposed gray artificial immunealgorithm can make more accurate prediction on the supply, demand and prices ofreal estate.