Deep learning has seen considerable success in several application sectors. Unfortunately, little research has been done on its efficacy in the context of network intrusion detection. This article includes case studies that use deep learning to identify network anomalies both supervised and unsupervised. It has been demonstrated that deep neural networks (DNNs) outperform current machine learning-based intrusion detection systems in the presence of shifting IP addresses. We also demonstrate how auto encoders can support network anomaly detection.
Keywords
Deep LearningIntruder Detection SystemDeep Neural NetworkthresholdAuto encoderAnomaly DetectionHost-based intruder
Subscribe to our newsletter
Sign up today to receive our latest research highlights, journals and event updates.
Our researchers are here to help
Ready to advance research with us?
Join a growing community of scholars and researchers, or submit your paper for peer-reviewed publication with Emirates Scholar.