Abstract
We propose a new oracle-based algorithm, BISTRO+, for the adversarial contextual bandit problem, where either contexts are drawn i.i.d. or the sequence of contexts is known a priori, but where the losses are picked adversarially. Our algorithm is computationally efficient, assuming access to an offline optimization 2 1 oracle, and enjoys a regret of order , where K is the number of actions, T is the number of iterations, and N is the number of baseline policies. 3 Our result is the first to break the barrier achieved by recent algorithms, which was left as a major open problem. Our analysis employs the recent relaxation framework of Rakhlin and Sridharan [7].