DeepDGA: Adversarially-Tuned Domain Generation and Detection

October 2016 Hyrum S. Anderson, Jonathan Woodbridge, Bobby Filar 9th ACM Workshop on Artificial Intelligence and Security (AISec)

#Adversarial ML #Detection

We use an adversarial training setup to generate domain names that evade existing DGA classifiers — and then use the same generator to train a more robust detector. One of the earliest applications of adversarial ML to a real network-security problem.

ACM AISec 2016 · DOI