Etat de l'art sur l'application des bandits multi-bras

TL;DR

This paper reviews recent advances in multi-armed bandit algorithms and their adaptations across various real-world applications, highlighting techniques like epsilon-greedy, UCB, and Thompson Sampling.

Contribution

It provides a comprehensive overview of how bandit algorithms are adapted for different practical scenarios and summarizes the current state of the art.

Findings

Various algorithms like epsilon-greedy, UCB, and Thompson Sampling are effectively adapted for different applications.

The review highlights the versatility of bandit algorithms in fields like clinical trials and network routing.

Recent results demonstrate improved exploration-exploitation balance in real-world problems.

Abstract

The Multi-armed bandit offer the advantage to learn and exploit the already learnt knowledge at the same time. This capability allows this approach to be applied in different domains, going from clinical trials where the goal is investigating the effects of different experimental treatments while minimizing patient losses, to adaptive routing where the goal is to minimize the delays in a network. This article provides a review of the recent results on applying bandit to real-life scenario and summarize the state of the art for each of these fields. Different techniques has been proposed to solve this problem setting, like epsilon-greedy, Upper confident bound (UCB) and Thompson Sampling (TS). We are showing here how this algorithms were adapted to solve the different problems of exploration exploitation.