Contextual Bandits Evolving Over Finite Time
Machine Learning
2019-11-15 v1 Machine Learning
Abstract
Contextual bandits have the same exploration-exploitation trade-off as standard multi-armed bandits. On adding positive externalities that decay with time, this problem becomes much more difficult as wrong decisions at the start are hard to recover from. We explore existing policies in this setting and highlight their biases towards the inherent reward matrix. We propose a rejection based policy that achieves a low regret irrespective of the structure of the reward probability matrix.
Cite
@article{arxiv.1911.05956,
title = {Contextual Bandits Evolving Over Finite Time},
author = {Harsh Deshpande and Vishal Jain and Sharayu Moharir},
journal= {arXiv preprint arXiv:1911.05956},
year = {2019}
}