ANOMALY DETECTION METHOD FOR NETWORK ATTACK WARNING

Abstract

In this study, the authors propose an anomaly detection method to warn of network attacks by using the Elastic Stack toolkit to collect and analyze log data of applications; then apply machine learning and PCA algorithm to detect behaviors, signs, anomalies in log data, thereby predicting user actions on applications that are attacks, invasions unauthorized entry or normal access; also compare the warning results of the proposed method with the Elastic machine learning technique being used in the Elastic Stack toolkit.