Intrusion detection in network systems through hybrid supervised and unsupervised mining process- a detailed case study on the ISCX benchmark dataset -
Saeid Soheily-Khah, Pierre-François Marteau, Nicolas Béchet · HAL (Le Centre pour la Communication Scientifique Directe) · 2017
Data mining techniques play an increasing role in the intrusion detection by analyzing network data and classifyingit as ’normal’ or ’intrusion’. In recent years, several data mining techniques such as supervised, semi-supervisedand unsupervised learning are widely used to enhance the intrusion detection. This work proposes a hybrid intrusiondetection (kM-RF) which outperforms in overall the alternative methods through the accuracy, detection rate, and falsealarm rate. A benchmark intrusion detection dataset (ISCX) is used to evaluate the efficiency of the kM-RF, and adeep analysis is conducted to study the impact of the importance of each feature defined in the pre-processing step.The results show the benefits of the proposed approach.