Achieving Near Instance-Optimality and Minimax-Optimality in Stochastic and Adversarial Linear Bandits Simultaneously
This work addresses the challenge of robust algorithm design for sequential decision-making in varying environments, representing a significant advance over prior state-of-the-art methods.
The paper tackles the problem of developing linear bandit algorithms that adapt to stochastic, corrupted, and adversarial environments, achieving nearly instance-optimal regret in stochastic settings and minimax-optimal regret in adversarial ones, with high-probability guarantees.
In this work, we develop linear bandit algorithms that automatically adapt to different environments. By plugging a novel loss estimator into the optimization problem that characterizes the instance-optimal strategy, our first algorithm not only achieves nearly instance-optimal regret in stochastic environments, but also works in corrupted environments with additional regret being the amount of corruption, while the state-of-the-art (Li et al., 2019) achieves neither instance-optimality nor the optimal dependence on the corruption amount. Moreover, by equipping this algorithm with an adversarial component and carefully-designed testings, our second algorithm additionally enjoys minimax-optimal regret in completely adversarial environments, which is the first of this kind to our knowledge. Finally, all our guarantees hold with high probability, while existing instance-optimal guarantees only hold in expectation.