Taming the Monster: A Fast and Simple Algorithm for Contextual Bandits

MSR-TR-2014-33 |

We present a new algorithm for the contextual bandit learning problem, where the learner repeatedly takes an action in response to the observed context, observing the reward only for that action. Our method assumes access to an oracle for solving cost-sensitive classification problems and achieves the statistically optimal regret guarantee with only ˜O(√T) oracle calls across all T rounds. By doing so, we obtain the most practical contextual bandit learning algorithm amongst approaches that work for general policy classes. We further conduct a proof-of-concept experiment which demonstrates the excellent computational and prediction performance of (an online variant of) our algorithm relative to several baselines.