Elimination algorithm for stochastic one-sided full-info bandits achieves O(sqrt(T log(TK))) distribution-independent regret and a gap-dependent bound, claimed as the best theoretical result.
In: Advances in Neural Information Process- ing Systems
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Stochastic One-Sided Full-Information Bandit
Elimination algorithm for stochastic one-sided full-info bandits achieves O(sqrt(T log(TK))) distribution-independent regret and a gap-dependent bound, claimed as the best theoretical result.