市场微结构的股市交易异常行为检测  

Anomaly Detection in Stock Marketplace Based on Market Microstructure

作　　者：林杨[1] 

机构地区：[1]福建师范大学经济学院,福建福州350117

出　　处：《福建师范大学学报（自然科学版）》2013年第1期31-35,67,共6页Journal of Fujian Normal University：Natural Science Edition

基　　金：国家自然科学基金资助项目(70371064)

摘　　要：股票市场存在诸多弊端,如滥用客户信息,价格操纵等.股市监控是金融监管体系中不可缺少的一环,它对市场交易的诚信、公平和公开透明起到重要作用.现有检测交易异常行为的诸多方法中,很少分析股市即日数据并挖掘潜在的交易行为来检测异常.股市是一个复杂的非线性系统,一套可行高效的异常行为检测方法是股市异常行为监控的重要课题.提出一种基于市场微结构的异常交易行为检测方法,该方法能较有效地检测出股市存在的异常交易行为.最后,通过实例说明该方法的可行性和有效性.It is well known that many defects exist in current stock market, such as intorma- tion abuse and price manipulation. Anomaly detection is helpful to enhance the integrity, fairness and transparence of stock market so it becomes a key link in financial regulatory system. Unfortu- nately, existing approaches were low performing as they rarely focused on analyzing the intraday in- formation and mining potential trading behaviors. It proposed a method, which based on market mi- crostructure, to detect abnormal trading behaviors. An experiment was presented demonstrating the feasibility and effectiveness of this approach.

关 键 词：市场微结构 即日数据 微结构向量序列 异常检测 

分 类 号：F830.91[经济管理—金融学]