Anomaly Detection of Financial Database Based on Navie Bayes Algorithm and k-means Clustering Algorithm
In order to effectively eliminate the security risks caused by user abnormal behavior to enterprise financial database,previous database anomaly detection technologies(such as Navie Bayes algorithm)usually use query feedback and establish user behavior characteristics(user behavior contour)to find security risks.However,this method takes more time to build training set and has no significant effect.Therefore,a financial database anomaly detection technology based on the combination of Navie Bayes algorithm and k-means clustering algorithm is proposed.By retrieving the user query contents and corresponding results in the operation log of financial database,k-means clustering algorithm is used to group users,and Navie Bayes algo-rithm is used to construct anomaly detection model.The test results show that compared with the traditional user behavior con-tour algorithm,the proposed algorithm has high training efficiency,greatly improved accuracy,improved comprehensive F1 value,and greatly improved the security of financial data.
万方数据
维普
