An Algorithm with Nearly Optimal Pseudo-Regret for Both Stochastic and Adversarial Bandits

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.

Cite

Text

Auer and Chiang. "An Algorithm with Nearly Optimal Pseudo-Regret for Both Stochastic and Adversarial Bandits." Annual Conference on Computational Learning Theory, 2016.

Markdown

[Auer and Chiang. "An Algorithm with Nearly Optimal Pseudo-Regret for Both Stochastic and Adversarial Bandits." Annual Conference on Computational Learning Theory, 2016.](https://mlanthology.org/colt/2016/auer2016colt-algorithm/)

BibTeX

@inproceedings{auer2016colt-algorithm,
  title     = {{An Algorithm with Nearly Optimal Pseudo-Regret for Both Stochastic and Adversarial Bandits}},
  author    = {Auer, Peter and Chiang, Chao-Kai},
  booktitle = {Annual Conference on Computational Learning Theory},
  year      = {2016},
  pages     = {116-120},
  url       = {https://mlanthology.org/colt/2016/auer2016colt-algorithm/}
}