The epoch-greedy algorithm for contextual multi-armed bandits

We present Epoch-Greedy, an algorithm for multi-armed bandits with observable side information. Epoch-Greedy has the following properties: No knowledge of a time horizon T is necessary. The regret incurred by Epoch-Greedy is controlled by a sample complexity bound for a hypothesis class. The regret scales as O(T2/3 S1/3) or better (sometimes, much better). Here S is the complexity term in a sample complexity bound for standard supervised learning.