An Improved PSO-Based Rule Extraction Algorithm for Intrusion Detection

The particle swarm optimization (PSO) algorithm is already proved efficient in the rule extraction in intrusion detection. But in practice the most intrusion detection systems often have a high false alarm rate. To solve it, this paper gives a new PSO-based algorithm which has a special fitness function to extract better rules set with lower false alarm rate to detect the attacks. Experiments based on the 1999 KDD cup data show that the algorithm can efficiently lower the false alarm rate.