On the price of exact truthfulness in incentive-compatible online learning with bandit feedback: A regret lower bound for WSU-UX

Open in new window