/CORe-Capitalizing-On-Rewards-in-Bandit-Exploration

This is the code for "CORe: Capitalizing On Rewards in Bandit Exploration" accepted to UAI'21

Primary LanguagePythonMIT LicenseMIT

CORe: Capitalizing On Rewards in Bandit Exploration

Demonstration codes of the paper "CORe: Capitalizing On Rewards in Bandit Exploration", which is accepted by UAI'21.

We apply CORe to both multi-armed bandit and linear bandit settings and compare it with several popular baselines.

If you find this repo useful, please cite our paper below.

@InProceedings{pmlr-v161-wang21c,
  title = 	 {CORe: Capitalizing On Rewards in Bandit Exploration},
  author =       {Wang, Nan and Kveton, Branislav and Karimzadehgan, Maryam},
  booktitle = 	 {Proceedings of the Thirty-Seventh Conference on Uncertainty in Artificial Intelligence},
  pages = 	 {1968--1978},
  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/wang21c/wang21c.pdf},
  url = 	 {https://proceedings.mlr.press/v161/wang21c.html},
}