Study On The City Residential Land Price Evaluation Based On Genetic Neural Network

With the development of Chinese urbanization, the impact of city land resources on the economic development is growing, at the same time, the effects of land price on land utilization regulation and land resources optimization and real estate market regulation are becoming increasingly evident. The city land price evaluation has been a method for the government and social entities who aim to obtain the land price information and make scientific decisions. As a regional land price, the benchmark land price has been directly effect on the different city land value, but the long update time and the simple correction method in the price evaluation, the city land price assessment accuracy will be affected. In addition, the other traditional land evaluation methods were restricted in global city land price evaluation, and could not take advantage of the land price affectors, so the author try to make research on the evaluation of city residential land prices from the view of artificial intelligence, which are a new method of land price evaluation.Based on the complex relationship simulation between the city residential land price and its influence factors, the paper take the grid technology, genetic neural network and other research methods to evaluate the city residential land price while combining theory and practice. The main research contents are as follows:firstly, based on the influencing factors system of residential land price combing, the paper came into being the influence factors; secondly, through the study of the genetic neural network theory and the analysis of the feasibility of the application of housing land price evaluation theory, the paper determined the construction ideas of evaluation model; thirdly, by determining the grid land price and influence factor scores and using MATLAB software to build evaluation model, the paper established the genetic neural network residential land price evaluation model and assessed the residential land price in Nanjing city.The conclusions of the paper are as follows:firstly, according to the city residential land factors classification system, the paper selected commercial factors, traffic factors, infrastructure and environmental factors; secondly, the genetic neural network has the capability of nonlinear mapping and global optimization, so the paper can simulate the relationship between land price and the influence factors, and finaly realize the land price evaluation; thirdly, taking Nanjing city as an example, the error on residential land price simulation is controlled within 2% while using the genetic neural network evaluation model based sample data, and by comparison with other model analysis the genetic neural network evaluation model is advanced and superiority;finaly, by the combination of the genetic neural netword land price evaluation model and the database of the Arcgis, we can not only realize the evaluation of the land price, but also realize the fast query.