摘要
采用神经网络模型研究信用风险评估问题,鉴于RBF神经网络计算量小、学习速度快,不易陷入局部极小而且具有很强的分类能力等优点,提出基于RBF神经网络的信用评估模型,通过试验该模型展现了良好的性能。
Adopts a neural network model to research on the problem of credit risk evaluation, in view of RBF neural network computing little, learning fast, not easy get into local minima, and it has a strong classifying capacity. So proposes a RBF neural network model has been proposed to assess the credit risk clients, this model shows a good performance through test in the end.
出处
《现代计算机》
2008年第10期154-156,162,共4页
Modern Computer
