Application Of BP Neural Network Model Combined With Grey Relation In Second Hand House Evaluation

Economic growth is the macro performance of industry development.In China,with the increase of the number and types of real estate transactions and related businesses,the real estate evaluation business is also rising.However,the more frequent real estate transactions are accompanied by some problems.As China's real estate evaluation is still in the exploratory stage,the three common evaluation methods in practice are inseparable from the subjective experience of professional evaluators,which makes the evaluation process not completely follow the objective basis,but mixed with some calculation results based on subjective judgment.In order to avoid this defect,the establishment of a quantitative evaluation method based on objective standards and reasonable factors can bring benefits to real estate evaluation in both theory and practice.Based on this purpose,this paper explores the real estate evaluation model with the help of artificial neural network theory in computer science.BP neural network model has strong learning ability and nonlinear mapping characteristics,which makes the model can deal with a large number of multi factor data well.The application of BP neural network model in second-hand housing price evaluation can not only achieve efficient processing of a large number of real estate transaction data information,but also avoid the subj ective influence of appraisers in traditional methods.The thesis reviews the related concepts,evaluation methods and research status of second-hand housing evaluation,and expounds the relevant theoretical basis of market comparison method,cost method and income method,as well as the advantages and disadvantages of various methods.In order to ensure the representativeness and consistency of the sample data,this paper uses the information of 450 second-hand housing transaction cases in 9 active administrative regions of Chongqing as the sample data when screening the influencing factors and training the BP neural network model.Secondly,in order to make the input BP neural network indicators are important factors affecting the price of second-hand housing,this paper uses the gray correlation analysis method to screen the factors affecting the price of real estate,and determines the most relevant nine factors.These factors need to be used as the input value sequence of BP neural network,and become the basis of building network model.Then,according to the characteristics of second-hand housing price evaluation,this paper constructs BP neural network model.Using the constructed network model,this paper uses MATLAB software to train the BP neural network model.The result shows that the error between the price of second-hand house evaluated by trained BP neural network and the real price is very small.After using the network model and market method to evaluate the same second-hand house,the results show that BP neural network model has more advantages in price evaluation,and it is an efficient and accurate evaluation method.