In this paper, we propose a novel solution to this problem by using generative adversarial networks to generate synthesized attack data for IDS. The synthesized attacks are merged with the original data to form the augmented dataset. Three popular machine learning techniques are trained on the augmented dataset. |