Privacy Leaks by Adversaries: Adversarial Iterations for Membership Inference Attack

Jing Xue, Zhishen Sun, Haishan Ye +4

Membership inference attack (MIA) has become one of the most widely used and effective methods for evaluating the privacy risks of machine learning models. These attacks aim to det…