摘要
我国现行反洗钱数据报告制度存在巨量数据、高误报率、缺乏自适应性和易于被规避等问题。针对这些问题,本文提出了基于机器学习的新一代数据监测与分析体系,通过识别客户交易行为模式正常与否来发现洗钱嫌疑行为,并对其体系框架、工作过程、技术可行性等方面进行了说明。该体系还可推广到预防金融欺诈和金融风险、客户关系管理等银行自身业务活动中。
The paper discussed the main problems of current anti-money laundering data reporting regulations, including enormous data, high false rate, lack of self-adaptive and easy evasion. A new generation of data monitoring and analytical framework that is based on machinery study is proposed to avoid such abuses effectively through the recognition of customer's behaviors pattern. This paper has made out its system frame, working course, technical feasibility, etc. The framework may also be applied into the other bank's businesses such as preventing from finance swindle and financial risks, customer relationship management.
出处
《中南财经政法大学学报》
CSSCI
北大核心
2005年第4期62-67,共6页
Journal of Zhongnan University of Economics and Law
基金
国家科技部"十五"攻关计划项目"商务智能在金融监管中的应用"(2004BA721A02)
关键词
反洗钱
数据监测
行为模式识别
金融监管
Anti-money Laundering
Data Monitoring
Behavior Pattern recognition
Financial Regulation