Comparison of anomaly detection techniques in networks

Anomaly detection in a network is important for diagnosing attacks or failures that affect the performance and security of a network. Lately, many anomaly detection techniques have been proposed for detecting attacks whose nature is strange. A process for extracting useful features is implemented in the anomaly detection framework. Standard matrices are applied for measuring the operation of the anomaly detection algorithms. This study compares different techniques for identifying anomalies which covers a wide spectrum of anomalies.