No Regrets for Learning the Prior in Bandits

Neural Information Processing Systems 

We consider a Bayesian version of the problem, where the bandit instances are drawn from some distribution.

Similar Docs  Excel Report  more

TitleSimilaritySource
None found