Detection of network anomaly based on hybrid intelligence techniques

Artificial Intelligence could make the use of Intrusion Detection Systems a lot easier than it is today. As always, the hardest thing with learning Artificial Intelligence systems is to make them learn the right things. This research focuses on finding out how to make an Intrusion Detection Systems...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:AL-Rafidain journal of computer sciences and mathematics Ročník 9; číslo 2; s. 81 - 98
Hlavní autori: Shahbaa I. Khaleel, Karam mohammed mahdi saleh
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Mosul University 04.12.2012
Predmet:
artificial intelligence
clustering algorithm
genetic algorithm
intrusion detection systems
swarm optimization
ISSN:1815-4816, 2311-7990
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Artificial Intelligence could make the use of Intrusion Detection Systems a lot easier than it is today. As always, the hardest thing with learning Artificial Intelligence systems is to make them learn the right things. This research focuses on finding out how to make an Intrusion Detection Systems environment learn the preferences and work practices of a security officer, In this research hybrid intelligence system is designed and developed for network intrusion detection, where the research was presented four methods for network anomaly detection using clustering technology and dependence on artificial intelligence techniques, which include a Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to develop and improve the performance of intrusion detection system. The first method implemented by applying traditional clustering algorithm of KM in a way Kmeans on KDDcup99 data to detect attacks, in the way the second hybrid clustering algorithm HCA method was used where the Kmeans been hybridized with GA. In the third method PSO has been used. Depending on the third method the fourth method Modified PSO (MPSO) has been developed, This was the best method among the four methods used in this research.
ISSN:1815-4816
2311-7990
DOI:10.33899/csmj.2012.163720