摘要
投资银行是金融体系的重要组成部分,从可量化监测的角度考察,选取了投资银行风险监测指标中包括市场风险、信用风险、流动性风险、资本风险在内的17个具体监测指标,构建了投资银行风险预警指标评价体系。运用基于数值优化的方法即L-M算法构建了基于前馈神经网络的投资银行风险预警模型。用训练好的BP神经网络模型,对检验样本进行了预测判别,结果显示出神经网络模型对我国投资银行风险具有较强的预测能力。
Investment Banks is an important part of the whole financial system. Grasps 17 analysis indexes which include market risk, loans risk, liquidity risk and capacity risk, and a new analysis indexes early warning system is put forward. Also, ajudging method of artificial neural network which is based on L-M algorithm is given in this paper. The application shows that the analysis indexes system and BP artificial neural network can evaluate the risk condition of investment bank effectively.
出处
《安徽工业大学学报（自然科学版）》
CAS
2006年第1期96-100,共5页
Journal of Anhui University of Technology（Natural Science）
关键词
投资银行
风险预警
BP神经网络
L-M算法
investment bank
early warning
BP artificial neural network
L-M algorithm
