1 paper · 1 filter
Minhao Cheng, Thong Le, Pin-Yu Chen +3
We study the problem of attacking a machine learning model in the hard-label black-box setting, where no model information is revealed except that the attacker can make queries to…