Distributed Thompson Sampling
Artificial Intelligence
2021-09-10 v2 Machine Learning
Abstract
We study a cooperative multi-agent multi-armed bandits with M agents and K arms. The goal of the agents is to minimized the cumulative regret. We adapt a traditional Thompson Sampling algoirthm under the distributed setting. However, with agent's ability to communicate, we note that communication may further reduce the upper bound of the regret for a distributed Thompson Sampling approach. To further improve the performance of distributed Thompson Sampling, we propose a distributed Elimination based Thompson Sampling algorithm that allow the agents to learn collaboratively. We analyse the algorithm under Bernoulli reward and derived a problem dependent upper bound on the cumulative regret.
Cite
@article{arxiv.2012.01789,
title = {Distributed Thompson Sampling},
author = {Jing Dong and Tan Li and Shaolei Ren and Linqi Song},
journal= {arXiv preprint arXiv:2012.01789},
year = {2021}
}
Comments
The paper is not finished and will not be updated