An Improved Relaxation for Oracle-Efficient Adversarial Contextual Bandits
Machine Learning
2023-11-13 v2 Data Structures and Algorithms
Machine Learning
Abstract
We present an oracle-efficient relaxation for the adversarial contextual bandits problem, where the contexts are sequentially drawn i.i.d from a known distribution and the cost sequence is chosen by an online adversary. Our algorithm has a regret bound of and makes at most calls per round to an offline optimization oracle, where denotes the number of actions, denotes the number of rounds and denotes the set of policies. This is the first result to improve the prior best bound of as obtained by Syrgkanis et al. at NeurIPS 2016, and the first to match the original bound of Langford and Zhang at NeurIPS 2007 which was obtained for the stochastic case.
Cite
@article{arxiv.2310.19025,
title = {An Improved Relaxation for Oracle-Efficient Adversarial Contextual Bandits},
author = {Kiarash Banihashem and MohammadTaghi Hajiaghayi and Suho Shin and Max Springer},
journal= {arXiv preprint arXiv:2310.19025},
year = {2023}
}
Comments
Appears in NeurIPS 2023