Machine Beats Machine: Machine Learning Models to Defend Against Adversarial Attacks

Jože M. Rožanec, Dimitrios Papamartzivanos, Entso Veliou +4

We propose using a two-layered deployment of machine learning models to prevent adversarial attacks. The first layer determines whether the data was tampered, while the second laye…