基于银行机构客户账户的可疑洗钱交易行为识别研究

《基于银行机构客户账户的可疑洗钱交易行为识别研究》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

1、西南交通大学硕士学位论文基于银行机构客户账户的可疑洗钱交易行为识别研究姓名：翁丽芳申请学位级别：硕士专业：管理科学与工程指导教师：高增安20061120西南交通大学硕士研究生学位论文第1I页AbstractThecrimeofMoneyLaundering(ML)canbeanindependentcriminalbehavior，andcanalsobepredicatecrimeandlowercrimeofothercriminalactivities．Therefore，MLwillnotonlyhave

2、animpactonthecountry’Seconomicdevelopment，butalsowillencouragecriminals，threatennationalstabilityanddoharmtonations’images．Withthedevelopmentofeconomicglobalizationandinternationalfinancialmarkets，Anti．moneyLaundering(AML)becomesanimportanttopicofinternationa

3、lfinancialfield．Meanwhile，SuspiciousActivityReport(SAR)systemisconsideredasthefoundationofAnti·moneyLaundering．BothinFinCenandChinaAnti—MoneyLaunderingMonitoring&AnalyzingCenter，theLargeandSuspiciousTransactionsusedforanalyzingaresuppliedbySARsystem．Thecurren

4、tSARsystemadoptedbyChinahassuchlimitationsasfuzzycriterions，lessofadaptiveabilities，easytoavoidandmassivedata，etc．Themainpurposeofthispaperistoputforwardanautomaticanddynamicsuspicioustransactionalactivitiesrecognitionmethodwhichisdifficulttoavoidandeasytomea

5、sure．AllofthesearebasedoncurrentSARsystemandbankaccounts．ThisisrathersignificantforthedevelopmentofSARsystemandaccordwiththefutureresearchdirectionofAML．Theup··to-·minuteresearchresultsofsuspicioustransactionactivitiesrecognitionathomeandabroadaregreatlyaccep

6、tedanddevelopedinthisthesis．BaseontheanalysisofcurrentSARsysteminChina，itdoesresearchOnsuspiciousactivitiesrecognitionviadatamining．Inthisprocess，OutlierDetectionTechnology，Grid—basedclusteringalgorithmandLOFalgorithmhavebeenapplied．Firstly，thecurrentAMLproce

7、duresinChinaareanalyzed．Atthesame西南交通大学硕士研究生学位论文第1II页time，itraisesthatSARsystemisthefoundationofAMLtoemphasizethesignificanceandvalueofthisstudy．ThenitanalyzesthecurrentSARsysteminChinaindetailespeciallythebadaspectsofittoindicatethenecessityofthisresearch．Th

8、eemphasisofthethesisliesinthatitputsforwardthealgorithmsmostsuitableforsuspiciousbehaviorrecognitionafteraseriesofcomparingamongseveralalgorithmsanddoessomeimprovement．What’Smoreimportant