Recognition: unknown
An algorithm with nearly optimal pseudo-regret for both stochastic and adversarial bandits
read the original abstract
We present an algorithm that achieves almost optimal pseudo-regret bounds against adversarial and stochastic bandits. Against adversarial bandits the pseudo-regret is $O(K\sqrt{n \log n})$ and against stochastic bandits the pseudo-regret is $O(\sum_i (\log n)/\Delta_i)$. We also show that no algorithm with $O(\log n)$ pseudo-regret against stochastic bandits can achieve $\tilde{O}(\sqrt{n})$ expected regret against adaptive adversarial bandits. This complements previous results of Bubeck and Slivkins (2012) that show $\tilde{O}(\sqrt{n})$ expected adversarial regret with $O((\log n)^2)$ stochastic pseudo-regret.
This paper has not been read by Pith yet.
Forward citations
Cited by 3 Pith papers
-
Spectral bandits
Spectral bandits achieve scalable regret in graph-structured recommendation by using an effective dimension to learn good policies from few node evaluations.
-
Budgeted Online Influence Maximization
A new algorithm for online influence maximization under a total budget constraint using the independent cascade model and edge-level semi-bandit feedback, with improved regret bounds for both budgeted and cardinality ...
-
Best of both worlds: Stochastic & adversarial best-arm identification
No algorithm can be optimal in both stochastic and adversarial best-arm identification; a new parameter-free algorithm matches the derived lower bound up to log factors in stochastic cases while handling adversarial rewards.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.