基于遗传神经网络的商业银行信用风险评估系统研究
Study of Risk Assessment of Commercial Banks Based on Genetic Neural Network
何泽恒 1朱虹1
作者信息
- 1. 哈尔滨商业大学 计算机与信息工程学院，黑龙江 哈尔滨 150028
- 折叠
摘要
近年来，商业银行信用风险的防范与管理一直备受各国重视。对其进行正确、有效的风险评估，应采用MATLAB工具和VisualC++程序构建基于遗传神经网络的商业银行信用风险评估模型，即运用遗传算法对BP神经网络进行优化，使模型实现优势互补，来仿真实现可操作的信用风险评估系统。以某商业银行提供的样本数据进行的实证分析结果表明：运用遗传算法优化权值的BP神经网络比未优化的BP神经网络对于商业银行信用风险的评估会更准确快速。这为制定我国商业银行信用风险的评估研究提供了参考，具有一定的理论和现实意义。
Abstract
In recent years, risk prevention and management of commercial banks have attracted the attention of various countries. Correct and effective risk assessment of commercial banks can be conducted by setting up a risk assessment model based on genetic neural net-works with the application of MATLAB instrument and Visual C++program, namely simulating the risk analysis system by optimizing BP neural networks with genetic algorithm that makes the models complement each other. The empirical analysis of sample data of a commer-cial bank proves that it is more accurate and speedy for commercial bank risk control to optimize the weight of BP neural networks by ge-netic algorithm than the networks without the optimization. The result offers a reference to the study of risk assessment of commercial banks in China.
关键词
商业银行/信用风险/BP神经网络/遗传算法/评估系统研究Key words
commercial bank/credit risk/BP neural network/genetic algorithm/assessment system study引用本文复制引用
出版年
2013
NETL
NSTL
维普
万方数据