Prediction Of Land Price Around Rail Transit Based On BP Neural Network

At the strong request of Fuzhou Government,Jiangxi Government and Nanchang Government initially agreed to co-invest in the construction of the intercity railway from Nanchang to Fuzhou.The investment of the provincial government is undertaken by Jiangxi Railway Investment Company as the main controlling party.According to the preliminary investment feasibility report,the project will lose money before 2035 and will basically achieve annual profit and loss balance after 2036.Generally speaking,this investment is not feasible.On the other hand,considering that the development of railway lines has multiple social significance,and the surrounding land price may change greatly with the development of railway projects,land becomes a possible way of compensation for project investment.As for the feasibility of this compensation method,it mainly depends on two factors,one is the land policy restriction,the other is the increase of land price.Therefore,for "Jiangxi Railway Investment",the future land price forecast becomes the key reference variable of whether the project can be invested and the corresponding equity arrangement.The main methods and achievements of land value assessment at home and abroad are sorted out,and the theoretical basis of land price is sorted out.In view of the particularity of land along urban rail transit,the feasibility and rationality of applying BP neural network to this kind of land valuation are considered,and the advantages and disadvantages of BP neural network in land valuation are discussed.In order to avoid the problems such as the sharp increase of network scale,the increase of operation time and the decrease of convergence and generalization ability of the network caused by the direct use of neural network in land price prediction and evaluation,the key comprehensive variables are obtained by using principal component analysis software in matlab.Secondly,the above key variables are applied to the BP neural network model to construct the BP neural network combined with the principal component analysis.The combined forecasting model evaluates the land value of the sample.By comparing the actual value with the predicted results of the model,it is verified that the model can effectively evaluate the value of land,and it also predicts the rising range of land price and land value around the station of Changfu Intercity Railway in the next ten to twenty years.The main conclusions are as follows: Although there is a major loss in the construction of Changfu Intercity Railway,the construction of Intercity Railway will bring about an increase in the surrounding land price.Combining with the forecast of the rising trend of the future land price around the site,if the average annual increase of the average land price of each site is about 4%,the loss can be compensated in general.Therefore,this investment is feasible.Through the analysis of the research examples in this paper,it is suggested that “Jiangxi Province Iron Investment” should fully integrate the resources of the surrounding land in thecontext of the policy of full use,strive for a more favorable range and area of comprehensive development land,and participate in the construction and industry of the surrounding areas along the site.Development;selection of hot spring stations,Nanchang South Station,with the greatest potential for appreciation,planning and comprehensive development according to the scale of 100 hectares,the remaining development sites of the remaining optional stations are 50 hectares,in order to compensate for the loss of the main investment.