Assessing Deviations of Empirical Measures for Temporal Network Anomaly Detection: An Exercise
The Internet and computer networks are exposed to an increasing number of security threats. With new types of attacks appearing continually, developing flexible and adaptive security oriented approaches is a severe challenge. In this paper, anomaly-based network detection techniques are a valuable technology to protect target systems and networks against malicious activities. However, despite the variety of such methods described in the literature in recent years, security tools incorporating anomaly detection functionalities are just starting to appear, and several important problems remain to be solved.