Analysis Of C Real Estate Data Firm’s Automatic Valuation Model

As China’s real estate industry develops, the demand of property valuation increases a lot and the traditional methods are hard to meet the increasing requirements of property transactions. With the development of Internet and computer technology, the automatic valuation techniques have been increasingly used in the real estate industry. Based on the real estate data of company C, this paper analyzed the accuracy rate and problems of automatic estimated model. On this basis, a series of suggestions were proposed.The contents of the article mainly talked about five aspects as follows. Firstly, company’s data collection and filtering methods were analyzed. Secondly, the article introduced the automatic real estate valuation models. Thirdly, this article compared the real data and estimated value of the automatic model and analyzed the effectiveness of the model. Fourthly, the problems of automatic valuation techniques and the defects of automatic valuation models were analyzed. Fifthly, recommendations for improvement of the model were proposed.Analysis showed that company C’s automatic valuation models need to take sample data as reference, because the quality of sample data directly affects the accuracy of assessment. What is more, data filtering method has to be perfected to improve the quality of data and it is also needed to exclude the problem data. Last but not least, a large number of samples should be taken as references in the models with high accuracy requirements.Based on research and analysis, this article proposed recommendations respectively on how to improve data accuracy, data filtering method and expand the amount of data samples. Among them, the issue suggested to do some research when doing the valuation and to figure out the difference between the sample data and the real market transaction data in order to solve the problem of data accuracy. For the data filtering method, this essay recommended to replace the traditional method with the ‘peripheral sampling’ way. For the expansion of sample data, this dissertation suggested to adopt the ‘sampling around the neighbor’ method to expand the amount of data.