UAI 2021poster14 citations

Variance-dependent best arm identification

Pinyan Lu, Chao Tao, Xiaojin Zhang

Abstract

We study the problem of identifying the best arm in a stochastic multi-armed bandit game. Given a set of $n$ arms indexed from $1$ to $n$, each arm $i$ is associated with an unknown reward distribution supported on $[0,1]$ with mean $\theta_i$ and variance $\sigma_i^2$. Assume $\theta_1 > \theta_2 \geq \cdots \geq\theta_n$. We propose an adaptive algorithm which explores the gaps and variances of the rewards of the arms and makes future decisions based on the gathered information using a novel approach called

BibTeX
@InProceedings{pmlr-v161-lu21a,
  title = 	 {Variance-dependent best arm identification},
  author =       {Lu, Pinyan and Tao, Chao and Zhang, Xiaojin},
  booktitle = 	 {Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence},
  pages = 	 {1120--1129},
  year = 	 {2021},
  editor = 	 {de Campos, Cassio and Maathuis, Marloes H.},
  volume = 	 {161},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {27--30 Jul},
  publisher =    {PMLR},
  pdf = 	 {https://proceedings.mlr.press/v161/lu21a/lu21a.pdf},
  url = 	 {https://proceedings.mlr.press/v161/lu21a.html},
  abstract = 	 {We study the problem of identifying the best arm in a stochastic multi-armed bandit game. Given a set of $n$ arms indexed from $1$ to $n$, each arm $i$ is associated with an unknown reward distribution supported on $[0,1]$ with mean $\theta_i$ and variance $\sigma_i^2$. Assume $\theta_1 > \theta_2 \geq \cdots \geq\theta_n$. We propose an adaptive algorithm which explores the gaps and variances of the rewards of the arms and makes future decisions based on the gathered information using a novel approach called
Variance-dependent best arm identification · UAI 2021