Network Intrusion Detection Based on Rough Set and K-nearest Neighbour

Increasing numbers of interconnected networks to the internet have led to an increase in cyber attacks which necessitates the need for an effective intrusion detection system. In this paper, two machine learning techniques: Rough Set (LEM2 Algorithm) and k-Nearest Neighbour (kNN) are used for intrusion detection. Rough set is a classic mathematical tool for feature extraction in a dataset which also generates explainable rules for intrusion detection. The experimental study is done on the international Knowledge Discovery and Data mining tools competition (KDD) dataset for benchmarking intrusion detection systems. In the entire experimentations, we compare the performance of Rough Set with k-Nearest Neighbour. The results generated from the experiment reveal that knearest neighbour has a better performance in terms of accuracy but consumes more memory and computational time. Rough Sets classifies at relative short time and employs simple explainable rules.