Open AccessHybrid deep-learning analysis for cyber anomaly detection
Author(s) -
Stanimir Kabaivanov,
Veneta Markovska
Publication year - 2020
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/878/1/012029
Subject(s) - anomaly detection , intrusion detection system , computer science , deep learning , artificial intelligence , anomaly (physics) , machine learning , anomaly based intrusion detection system , data mining , warning system , real time computing , telecommunications , physics , condensed matter physics
Cyber threats evolve continuously and so do the detection tools and algorithms. In this paper we analyse the efficiency of hybrid deep-learning analysis as a mean to detect anomalies in computer network traffic. Different deep-learning algorithms are tested against real network intrusion events in an attempt to assess their potential as an early warning system. We suggest a combination of algorithms and rule-based filters as a hybrid system that can improve efficiency and accuracy of cyber anomaly detection.