摘要
商业银行风险由市场风险、信用风险和操作风险三类组成。随着金融产品不断创新,商业银行交易操作变得日益复杂,由于交易风险引发的商业银行安全事故频出。目前,商业银行主要从规章制度上规范交易操作步骤以及使用信息系统对当天交易产生的票据进行复核做到事后监督,而对正在操作中的交易只能通过业务部门制定的规则对交易报文进行规则匹配,但也只能处理已知的风险交易。为了保证国家金融市场秩序的稳定和商业银行正常的经营,迫切需要一种能有效识别出商业银行日常操作交易中关于风险交易的方法,针对这些问题,在整合商业银行风险交易和数据挖掘技术的基础上,设计了基于交易报文指纹关系的风险交易识别方案。
Commercial banks have market risk,credit risk and operational risk of three categories. As the financial products innovation,commercial banks trading operations are becoming increasingly complex,due to the risk caused by lots of commercial banks to a security incident. Currently,the main commercial banks used trading rules and regulations on specification operation steps and information systems audit to check the day trading paper post-supervision,and also dealed the only through business rules for trading message rules matching,but also could deal with known risk trades. In order to guarantee the stability of national financial market order and normal operation,commercial banks urgently need an effective method which identify the daily operation of commercial bank deal about risk trade. To solve these problems,based on the integration of the commercial bank risk trade and data mining technology,it designed a risk identification scheme based on the transaction message fingerprint relationship.
出处
《计算机应用研究》
CSCD
北大核心
2017年第3期800-804,共5页
Application Research of Computers
关键词
商业银行
风险交易
机器学习
指纹特征
commercial banks
risk trades
machine learning
fingerprint characteristics
