摘要
本文选取了近千家企业的财务数据作为样本,建立了基于BP神经网络的企业财务风险评价模型,通过对实验样本的抽取以及标准化处理,进行了模型优化,选择了适合解决该问题的网络参数,并利用试错法找到了误差性能最好的网络结构,通过网络测试,对该模型的可靠性进行了分析。结果表明,该模型优于传统的专家评价方法。
This paper collects the financial date from nearly a thousand enterprises as samples and builds up and enterprise financial risk evaluation model based on BP neural network. This paper carries out the model optimization through sampling and standardization and finds out the best error performance network structure using the method of trial-and-error and also conducts the stability analysis of the model through network testing. The result shows that the model is superior to the traditional expert assessment method.
出处
《电脑与电信》
2006年第8期12-17,共6页
Computer & Telecommunication
关键词
BP神经网络
风险评价
指标体系
标准化
back propagation neural network
risk evaluation
index system
standardization