Robust Stochastic Linear Contextual Bandits Under Adversarial Attacks
This addresses a critical security problem for applications like recommender systems and online advertising by providing a robust solution in the contextual bandit setting, which is a significant advancement over prior non-contextual methods.
The paper tackles the vulnerability of stochastic linear contextual bandit algorithms to adversarial attacks, which can cause complete failure, by proposing the first robust algorithm that works under fully adaptive and omniscient attacks on both rewards and context, achieving sub-linear regret without needing attack details. Experimental results show improved robustness against various popular attacks.
Stochastic linear contextual bandit algorithms have substantial applications in practice, such as recommender systems, online advertising, clinical trials, etc. Recent works show that optimal bandit algorithms are vulnerable to adversarial attacks and can fail completely in the presence of attacks. Existing robust bandit algorithms only work for the non-contextual setting under the attack of rewards and cannot improve the robustness in the general and popular contextual bandit environment. In addition, none of the existing methods can defend against attacked context. In this work, we provide the first robust bandit algorithm for stochastic linear contextual bandit setting under a fully adaptive and omniscient attack with sub-linear regret. Our algorithm not only works under the attack of rewards, but also under attacked context. Moreover, it does not need any information about the attack budget or the particular form of the attack. We provide theoretical guarantees for our proposed algorithm and show by experiments that our proposed algorithm improves the robustness against various kinds of popular attacks.