Adapting to Delays and Data in Adversarial Multi-Armed Bandits

                      Andras Gyorgy 1 Pooria Joulani 1

            Abstract                1. Introduction
                                The multi-armed bandit problem is a canonical model for
   We consider the adversarial multi-armed bandit      sequential decision making with limited feedback. In this
   problem under delayed feedback. We analyze        model a learner makes a sequence of actions. After ev-
   variants of th ...