Research And Implementation Of Securities Trading Abnormal Behavior Detection Based On Data Mining

At present,with the increasing popularity of the Internet,securities trading based on the Internet is becoming more and more popular.And the bad influence of the abnormal behaviors in the securities trading process is also increasing.Therefore,it is essential to find and detect the various kinds of abnormal behavior in the securities trading system to maintain the Internet securities transaction.It is of great significance to develop a securities trading system with anomaly detection function to protect the benefit of users.This dissertation mainly focuses on the detection of abnormal behavior in the securities trading system.The corresponding algorithm is proposed and tested.Moreover,the proposed algorithm is integrated into the securities trading system.The securities trading system is designed and implemented with the function of abnormal behavior detection.The structure of this dissertation is as follows:the first chapter mainly introduces the current situation of the securities transaction,research background and practical significance;the second chapter introduces the commonly used securities analysis techniques and data mining techniques;the third chapter analyses the requirement of abnormal behavior in the securities trading detection technology,and analyses the feasibility from three aspects of technology,economics and operation;the fourth chapter studies the K-means cluster algorithm with adaptive ability to detect abnormal behavior,and carries out the related test and corresponding result;the fifth chapter develops and tests the securities trading system with abnormal behavior detection.