Decentralized Adversarial Training over Graphs

Ying Cao, Elsa Rizk, Stefan Vlaski +1

The vulnerability of machine learning models to adversarial attacks has been attracting considerable attention in recent years. Most existing studies focus on the behavior of stand…