Warm-starting Contextual Bandits: Robustly Combining Supervised and Bandit Feedback.

Chicheng Zhang, Alekh Agarwal, Hal Daum III, John Langford, Sahand Negahban

Browse the full ICML paper archive.