基于支持向量机的商业银行信用风险评估

Credit Risk Assessment in Commercial Banks Based on SVM

支持向量机(SupportVectorMachine,简称SVM)是在经验风险最小化原理上发展出的一种新的机器学习技术,在解决小样本、非线性及高维模式识别问题中表现出许多特有的优势。论文首先详细介绍了支持向量机的线性和非线性分类算法,然后将支持向量机非线性分类器应用于银行信用风险的评估中,最后分析对比了选用不同核函数和参数的实验结果。

Support Vector Machines(SVM)is a novel machine learning technique developed on empirical risk minimi-zation principle.SVM has many advantage in solving small sample size,nonlinear and high dimensional pattern recognition problem.Firstly SVM linear and nonlinear classifiers algorithm are discussed detaily in the paper.Then SVM nonlinear classifiers is applicated in Credit Risk Assessment of commercial Banks.At last this paper analyses and compares the experiment results by selecting different knernel function and parameters.