Application Of Artificial Neural Network In The Residential Real Estate Appraisal

With the development of market economy, the real estate transaction hasbecome increasingly dynamic.Real estate transfer, lease, mortgage, taxes andother economic behaviors based on its value are happening in everyday. Thereal estate appraisal is an important guarantee of real estate pricesnormalization and fair trade and a powerful driving force for theestablishment of a healthy and orderly real estate transactions. However thereare still many problems in real estate appraisal industry,for example, theresults of the assessment too dependent on the experience of theappraisers,the lack of accurate assessment base and difficult to guarantee thequality of assessment, these problems have hindered the development of realestate appraisal. At present, the research on the real estate appraisal methodshas become a hot social issues.This paper studies the market method using in the residential real estateappraisal. Firstly it will introduce to the significance of the real estateappraisal and the station of research and the theory and methods of real estateappraisal and establish a index set of elements of real estate price,then revealthe shortcomings of the traditional marketing methods;The writer introduceartificial neural network method based on MATLAB software in this paper inorder to conquer the shortcomings effectively, This article base on30sold realestate information, selecting the parameters through trial and error,includinghidden layer nodes and the algorithm.other parameters selected base ontheory. then,applying the optimal parameters selected to the model.Theresults show the model has excellent simulation ability, the convergence speed,high precision.The writer also introduce wavelet analysis theory to optimizedBP neural network,raising the idea of wavelet neural network for real estateassessment and establish the model,The results show that the model is betterthan BP neural network model in simulation ability.This article establish a scientific evaluation index system and combinereal estate appraisal theory with information technology together perfectly.To a great extent, it reduces the influence of human factors on the accuracy ofassessment results and is a study of great practical significance and broadmarket prospects.