The Problems Studied Based On Support Vector Machine (SVM) Method Of Small And Medium-Sized Enterprise For Credit Rating

Small and medium-sized enterprises are at an important strategic position in the national economy and social development, but they have been faced with financing difficulties for a long time, restricting its healthy development. The reason of difficulty in financing is various, but it will largely solve the information asymmetry problem between banks and small and medium-sized enterprises and relieve its financing difficulties if we could build a set of small and medium-sized enterprise characteristics for the credit rating method, However, China’s small and medium-sized enterprises’ credit rating method largely baized on enterprise at present. Dung to its rating results do not reflect true credit level accurately, it is difficult to reflect the small and medium-sized enterprise credit risk. In recent years, small and medium-sized enterprise credit rating research is in the ascendant, gained a series of achievements, and has accumulated a lot of successful experience, but it still lacks of a set of characteristics which reflecting small and medium-sized enterprises’own special credit rating system. Then we need to optimize the small and medium-sized enterprises’credit evaluation index system and the technical route, providing financial support service for the healthy development of small and medium-sized enterprises.This research focuses on selecting small and medium-sized enterprises’credit rating index system and technical route optimization. At first, this article analysis and research the credit ratings related theory detailed, and then go through to the defect of commercial Banks index system, according to their own characteristics of small and medium-sized enterprises and the level of credit status system analysis. At last, we established a set of small and medium-sized enterprises on the basis of their own credit rating index system. In order to overcome limitations of the existing traditional statistical model based on the rating method, this study tries to convert credit rating to pattern recognition and cluster analysis, By selecting a small sample learning theory support vector machine (SVM) method, we use this to assess small and medium-sized enterprise credit. Then we gained advanced credit rating method relatively of small and medium-sized enterprise. This paper has introduced this method in details at first, then through empirical analysis and comparing to BP neural network proved the effectiveness of this method, and intended the future research at last.