The Price Evaluation Research Of Second-hand House Based On The Random Forest Model

With the development of economy, economic activities related to real estate are becoming more frequent. Because of the less land can be exploited, the second-hand housing transactions are becoming more active. The need for real estate appraisal is increasing. Weather from the view of market participants or the tax collector, how to assess the second-hand housing is an eternal theme. At present, the use of traditional methods including comparison method and cost method and income method mainly depend on the experience of appraisers, rather than the mathematics statistics models, so it will costs more in real life.In this dissertation, Hedonic pricing model has been introduced to understand the price of second-hand house from the demand of customer. In the process of building second-hand house evaluation model, major authors employ multiple linear regression method, who made the assumption of linearity and caused greater errors. Random Forest, an method from machine learning theory, has been introduced to make an evaluation model of second-hand housing in the thesis.Finally, the model has been used to assess a certain area of Guangzhou city. By adopting the listing data of second-hand housing and method of visiting residence community, author got298items about characteristic vectors in the study area to build the empirical model and make a prediction. Author used the traditional multiple linear regression method to make a comparison with the model built above. The evaluation results of two methods show that the second-hand housing pricing evaluation model based on Random Forest has the better accuracy, which is worth for spreading application.