Anomaly detection based on profile signature in network using machine learning technique

Abstract

Within the last couple of years, a lot of IDSs have already been developed, both as research prototypes and as commercial systems. The aim of Intrusion Detection would be to positively identify all true attacks and adversely identify all non-attacks. Through the study of the procedure and signature of intrusion behaviors, IDS can produce a real-time reaction to intrusion occasions and invasion processes. The objective of this study is to work with appropriate algorithms for intrusion detection and hybrid them to find acceptable results and investigate on new hybrid techniques of intrusion detecting system with high accuracy. Moreover, it also aims to minimize a false positive and false negative alarms in networks.