The Commercial Banks Housing Credit Assessment Model Is Set Up By Using The SVM

With the development of housing marketing in our country, the scale of consumer housing loans of commercial bank is increasing rapidly. At the same time, an intense rise in loan risks of commercial bank is witnessed and the bad loan ratio in our country is obviously higher than that in the developed countries. Therefore, it's very important for commercial bank to solve the risk problem in the development of housing loans, as well as establish an effective risk prevention system.Based on the above considerations, in this paper, the commercial banks housing credit assessment model is set up by using the SVM (Support Vector Machine) model developed and widely applied in recent years. Then a commercial bank's housing credit data is used for applied research. This paper analyzes the impact of China's housing credit default factors, combining housing credit system to assess the problems, proposes a suitable housing credit assessment indicators system from the aspects of housing characteristics, characteristics of the borrower, loans and derivatives of variable characteristics. A SVM model based on the posterior probability is bring forward by using a commercial bank's housing credit data. This model can give the posteriori probability of customers'default. According to the posteriori probability, we can divide customer's credit rating of housing credit, so as to formulate the corresponding housing credit policies. Compared to the complied results of BP neural network and logit model, SVM model based on the posterior probability has a very high classification accuracy and solidity. Thus this model has better applicability for commercial banks to control housing credit risk.