基于BP模型的商业银行贷款风险预测  

Forecast of Commercial Bank Loaning Risk Based on BP Model

作　　者：李家军[1] 李娅娅[1] 

机构地区：[1]西北工业大学经济系,陕西西安710072

出　　处：《计算机仿真》2008年第3期278-280,326,共4页Computer Simulation

基　　金：国家社科基金资助项目(05BJY021);西安市社会科学基金项目(06J52)

摘　　要：国有商业银行不良贷款严重束缚了商业银行的发展。防范金融风险,降低不良贷款,增强商业银行的风险识别能力,其关键是风险预测。商业银行贷款本身是一个复杂的非线性系统,用一般的线性理论难以客观反映其规律,为此,采用人工神经网络方法进行研究。在简述概念的基础上,通过反向传播网络(BP模型)对贷款企业的经营能力进行预测,从而对商业银行的贷款决策提供理论支持,使商业银行贷款风险能被控制在可控范围内。并对代表性的A股份有限公司进行预测,所得结果表明:建立的预测模型具有良好的风险预测能力。In the field of the prevention of financial risks, it is the key to forecast the risks of commercial banks. Commercial bank loaning is a complex and nonlinear system, and it is difficult for the linear theory to reflect the general laws, thus this paper adopts artificial neural network to research them. In virtue of back - propagation network （BP） , the operating capacity of the loan can be forecasted, therefore the commercial bank loaning risk will be within a controllable range. The results through forecasting a typical company as a representative reveal that the established prediction model is capable of predicting risk.

关 键 词：人工神经网络 反向传播模型 风险 预测 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]