Applicative Study Of Artificial Neural Network In Real Estate Appraisal

With the rapid development of economy and the improvement of the people's living conditions, more and more people have desire to improve their housing conditions, which induces the blossom of real estate market and the increase of marketing quantity, and even the rapid development of the appraisal of real estate. The traditional real estate appraisal approach is that real estate appraisal personnel make an estimate of the real estate price according to the appraisal procedure combining their appraisal experience, and even on the basis of analyzing all kinds of factors that may influence the real estate price.Because of the complex characteristic of real estate appraisal, such as actualization characteristic, market characteristic, and forecast characteristic, the traditional real estate appraisal depend too much on appraisal experience of appraisal personnel, which make the appraisal result difficult to guarantee the objectiveness, science, and justice. Considering the complexity, uncertainty, and even nonlinear property between the real estate price and its influence factor, artificial neural network is introduced to improve the appraisal approach in this paper. This method can reduce the deviation of appraisal result due to the subjective factor of the appraisal personnel. With the help artificial neural network, the appraisal personnel can provide a rapid, accurate and fair real estate appraisal. On the basis of large quantity of former research, the main works of this paper is shown below:1. Several general used real estate appraisal approaches are introduced, and principle and characteristic of real estate appraisal are analyzed too. The relate theory of artificial neural network and even its arithmetic model are sorted and analyzed. Furthermore, market comparison approach is selected as an example to improve the appraisal approaches.2. Feasibility of artificial neural network used in the real estate appraisal is demonstrated. And even the principle and process of improvement of this method are introduced. On the basis of the BP artificial neural network, an appraisal model is provided. With this model, the real estate price influence index system is set up. Finally, this index system is quantified to make it suit to the training input data of the artificial neural network.3. Based on the appraisal model provided above, the design of real estate appraisal system is given in this paper. Furthermore, a program is compiled with VC++ language to improve the computer appraisal approach and to enhance the work efficiency of appraisal personnel.4. The problems encountered in the course of this research and even the deficiency of this design system are summarized. And the subsequent improvement idea is given to provide a reference for the later researchers.