Action-Manipulation Attacks on Stochastic Bandits

Guanlin Liu, Lifeng Lai

Abstract

As stochastic multi-armed bandit model has many important applications, understanding the impact of adversarial attacks on this model is essential for the safe applications of this model. In this paper, we propose a new class of attack named action-manipulation attack, where an adversary can change the action signal selected by the user. We investigate the attack against a very popular and widely used bandit algorithm: Upper Confidence Bound (UCB) algorithm. Without knowledge of mean rewards of arms, our proposed attack scheme can force the user to pull a target arm very frequently by spending only logarithm cost.
BibTeX
@inproceedings{liu2020action,
  author = {Guanlin Liu and Lifeng Lai},
  title = {{Action-Manipulation Attacks on Stochastic Bandits}},
  booktitle = {IEEE International Conference on Acoustics, Speech and Signal Processing},
  year = {2020},
  doi = {10.1109/icassp40776.2020.9053495},
}

Loading related content...

Figure