Deep IDS : A deep learning approach for Intrusion detection based on IDS 2018
Arunavo Dey
Abstract
Intrusion Detection is one of the fields network security important for industry 4.0. Applying deep learning models opened a new scope in this field. But availability of latest data set and volume makes it often harder to apply latest techniques. Moreover emergence of new machine learning algorithms always hold scope to improve over the existing ones. In this paper, the effectiveness of attention mechanism over the existing deep learning techniques for Intrusion detection is being proposed and a novel attention based CNN-LSTM model has been proposed based on IDS 2018 data set. A detail performance evaluation on IDS 2018 has been elaborated to establish the claim.
Topics & Concepts
Computer scienceIntrusion detection systemDeep learningArtificial intelligenceScope (computer science)Machine learningField (mathematics)Set (abstract data type)Network securityData setData miningComputer securityMathematicsPure mathematicsProgramming languageNetwork Security and Intrusion DetectionAnomaly Detection Techniques and ApplicationsAdvanced Malware Detection Techniques