Theimproved SOM Based Financial Analysis And Early Warning Research Of Real Estate Enterprises

In recent years, the Real Estate Industry is developing rapidly.With the fast development of this industry, the industry concentration degree is raising, and the number of the Real Estate development projects is growing. The development of China’s Real Estate Industry and the improvement of the Real Estate policy, financial policy and innovation, continue to create more and more space of development for the Real Estate Enterprises. However, as a pillar of our national economy, the Real Estate Industry has the characteristics of capital intensive, long investment return period and big risk. At the same time, China’s Real Estate Industry is still not standardized and the financial system is not perfect. So the Real Estate Industry is bearing the enormous financial risk with the rapid development. Neural network method is robust, non parametric and has nonlinear mapping ability. Its learning experience is strong and it has high classification accuracy. The SOM(Self-Organizing Feature Map Network) model is a non parametric clustering methods.It’s unsupervised and it has strong adapting ability and good clustering function. This study mainly studies on how to evaluate the financial risk of enterprises with the actual application of SOM network. The SOM was improved to improve the accuracy of the solution for the research of financial early warning of the Real Estate Enterprises in the use of the improved SOM. And the results were compared and summarized with the financial early warning method of other Real Estate Enterprises. It can be used for guidance and applied to practice.This article is based on the current situation of financial risk of Real Estate Enterprises which are list on the Shanghai and Shenzhen stock market. It uses 94 Real Estate Enterprises and 28 financial early warning indexes to construct the Real Estate financial risk evaluation model. First, this article interprets each part of the financial risks of China Real Estate Enterprises of the risk through the analysis. So the main problem at the present stage is analysising the causes and influencing factors. Second, it establishes SOM neural network clustering model. Third, it sets up the indexes to evaluate the financial risk of the Real Estate business model. Then it collects Shanghai and Shenzhen Real Estate listing Corporation’s financial information to finish the 94 Real Estate listing Corporation’s financial statements. And it selected 28 financial indicators of the specific data. Fourth, the SOM neural network has been improved to improve the accuracy of the solution. The improved SOM is used on 94 real estate listing Corporation financial data sample to make clustering analysis. The results make an integrated clustering precision on the financial characteristics of each enterprise. Then it makes compare and summary with other enterprises financial early warning method. Fifth, in view of the above method of the results of the Real Estate Enterprises, it gives financial risk management and financial early warning advices.