Adversarial attacks can cause DNS amplification fool network defense systems machine learning study finds


The application of machine learning and deep learning in cybersecurity, particularly in network intrusion detection and prevention, has attracted increasing attention in recent years.

Researchers at the Citadel, a military academy in South Carolina, the US, found that adversarial attacks�specially produced data that tricks neural networks into changing their behavior�can be used to circumvent deep learning models trained for network intrusion detection. Read More…