Pricing Of Housing Reverse Mortgages Real Estate Appraisal Based On BP NN Model

During the course of the continuous increasing elderly population, the aged mode is also changing. The emergence and introduction of "House-For-Pension" pattern has great significance to relieve the aging crisisIn this paper, by using comparative research methods, theoretical analysis and research for reference method, system analysis and research methods and related research methods, from the plight of the pension in our country at present, research on the present housing reverse mortgage loan the new endowment mode of theoretical significance and practical significance.With housing endowment the success of the product design, and is based on house value evaluation. Retirement with house, once successful, will need to be able to realize the more efficient and exact mass appraisal of real estate appraisal method. In this paper, using the MATLAB neural network toolbox, combining with the real estate appraisal method of market comparison method, adopt the method of simulation training, the BP neural network evaluation model is established. In specific operations, first of all, through the chain of home real estate, second-hand housing transactions institutions such as soufun looking for comparable cases, and then through the expert scoring method, select11important indicators and a quantitative influence the real estate price. In this paper, a total of35cases of second-hand housing transaction case.In35cases,1to30as training samples, training to simulate neurons. Neural network toolbox can carry on the automatic matching, approximating function to the input data. After repeated training and iterative error analysis, until the error is reduced to stop training, desired level at this point, affect the nonlinear function relationship between the basic form of the price of the house. Followed by31to36cases trading as test samples, test the accuracy of the model. The results showed that the BP neural network was used to simulate the stock model accuracy is higher. Finally to quantify the cases to estimate the properties of related indicators in the input model, relative reliable to appraise the real estate market value.