Learning Strategy-Aware Linear Classifiers

Yiling Chen, Yang Liu, Chara Podimata

We address the question of repeatedly learning linear classifiers against agents who are strategically trying to game the deployed classifiers, and we use the Stackelberg regret to…