Characterizing the Optimal $0-1$ Loss for Multi-class Classification with a Test-time Attacker

Characterizing the Optimal

0-1

Loss for Multi-class Classification with a Test-time Attacker

2023

NeurIPS