Designing a holistic end-to-end intelligent network analysis and security platform

• Published in: Journal of physics. Conference series Vol. 978; no. 1; pp. 12100 - 12108
• Main Author: Alzahrani, M
• Format: Journal Article
• Language: English
• Published: Bristol IOP Publishing 01.03.2018
• Subjects: Deep learning; Fuzzy logic; Intelligent networks; Machine learning; Malware; Network analysis; Neural networks; Physics
• ISSN: 1742-6588; 1742-6596
• DOI: 10.1088/1742-6596/978/1/012100

Firewall protects a network from outside attacks, however, once an attack entering a network, it is difficult to detect. Recent significance accidents happened. i.e.: millions of Yahoo email account were stolen and crucial data from institutions are held for ransom. Within two year Yahoo's system administrators were not aware that there are intruder inside the network. This happened due to the lack of intelligent tools to monitor user behaviour in internal network. This paper discusses a design of an intelligent anomaly/malware detection system with proper proactive actions. The aim is to equip the system administrator with a proper tool to battle the insider attackers. The proposed system adopts machine learning to analyse user's behaviour through the runtime behaviour of each node in the network. The machine learning techniques include: deep learning, evolving machine learning perceptron, hybrid of Neural Network and Fuzzy, as well as predictive memory techniques. The proposed system is expanded to deal with larger network using agent techniques.