Reinforcement Learning for Markovian Bandits: Is Posterior Sampling more Scalable than Optimism?

Re-assign community