Commercial Banks Internal Credit Rating And Application Research On Support Vector Machine

Risk facing china commercial banks is still credit risk mainly. It's particularly important for domestic commercial banks to improve their risk management ability when competing globally. Credit rating system is the fundamental of risk management for commercial banks, and supports the operation of credit businesses as well. With the changing of global finance competing circumstance, requirements of precisely credit risk measuring, evaluating of risk adjusted return and capital adequacy, all bring forward request to inovate the fundamental credit rating system. With the rapidly advancing of information techlonogy, an aspect for future research is about how to build machine learning and artificial intelligence approaches based credit rating system.The paper focuses on theories and problems concerning commercial banks'internal rating in detail, and analyzes issues when applying statistical approaches to credit rating too.Then the paper studies most of the classical models of pattern recognition and analyzes the applicability of these models for rating problem.Furthermore, the paper studies theories and algorithm of statistical learning theory and support vector machine, and shows its advantages over other models on generalization ability. At last, using data of listed companies, the paper gives a positive example of the effect when applying SVM to credit analyzing problem. The paper concludes support vector machine is suitable to process credit rating problems.