Optimality of Thompson Sampling with Noninformative Priors for Pareto Bandits

Open in new window