The intrusion detection system based on a novel association rule
Baoping Gu, Honyan Guo · 2014
Because of its accurate and robust performance, association rule algorithm is recently used for intrusion detection. However, the existing algorithms for associative classification suffer from inefficiency: high misinformation rate and low detection rate, addressing this problem, a novel association rule is presented and successfully used in intrusion detection. Mining only the atomic association rules achieves fast intrusion detection classification. Using the strong atomic association rules, extracted under a high confidence threshold, multiple passes of partial classifications can classify the whole dataset. This algorithm uses a self-adaptive confidence threshold and a dynamic support threshold. The experiments were performed on a standard dataset of KDD cup99. The results show the proposed algorithm can systematically keep low condition of misuse rates, intrusion detecting rates improve to some extent.