Robert Schapire

Latest

• On the Hardness of Bandit Learning
• Provable Interactive Learning with Hindsight Instruction Feedback
• A Unified Model and Dimension for Interactive Estimation