摘要
对信用风险、信用评分进行了分析,在综合分析国内外企业信用评分指标体系的基础上,结合我国企业信用评分的特点,建立了适合我国企业信用评价的指标体系。结合国内外相关研究的现状与进展,及信用评分本身所具有的特点,建立了基于径向基函数神经网络的信用评分模型,利用现有数据分别进行判别和分析,研究其计算结果与实际情况的差距,然后使用改进的RBFNN学习算法,对径向基函数神经网络进行了学习训练,得到了令人满意的评价结果。利用该模型建立的评分系统具有进一步研究和推广应用的价值。
Firstly analyze the credit risk and credit scoring. After comprehensive analysis of the indicator systems of both domestic and abroad enterprises,and combining with the character of China enterprise,built the indicator system suitable for China. Combining with the current status and the development of related studies and the properties of credit scoring itself, built the credit scoring model based on the radial basis function neural network. Utilizing the available data, carried out the differentiation and analysis separately, studied the difference between the calculated results and the practical situation, then used the improved RBFNN train algorithm to train the radial basis function neural network, and obtained the satisfactory results. The credit scoring system using the model is worthy to study and popularize further.
出处
《计算机技术与发展》
2007年第9期11-14,共4页
Computer Technology and Development
关键词
径向基函数
神经网络
信用评分
指标体系
radial basis function
neural network
credit score
indicator system
