Akshay Krishnamurthy, John Langford, Aleksandrs Slivkins, Chicheng Zhang.
Year: 2020, Volume: 21, Issue: 137, Pages: 1−45
We study contextual bandit learning with an abstract policy class and continuous action space. We obtain two qualitatively different regret bounds: one competes with a smoothed version of the policy class under no continuity assumptions, while the other requires standard Lipschitz assumptions. Both bounds exhibit data-dependent “zooming” behavior and, with no tuning, yield improved guarantees for benign problems. We also study adapting to unknown smoothness parameters, establishing a price-of-adaptivity and deriving optimal adaptive algorithms that require no additional information.