Towards Optimal Algorithms for Multi-Player Bandits without Collision Sensing Information

TL;DR

This paper introduces a new algorithm for multi-player multi-armed bandits that operates without collision sensing and does not require prior knowledge of minimal expected rewards, achieving better theoretical and practical performance.

Contribution

The proposed algorithm removes the need for a lower bound on minimal expected rewards and improves scalability, with proven regret bounds and superior empirical results.

Findings

Outperforms existing algorithms in simulations

Does not require prior minimal reward bounds

Achieves lower regret bounds

Abstract

We propose a novel algorithm for multi-player multi-armed bandits without collision sensing information. Our algorithm circumvents two problems shared by all state-of-the-art algorithms: it does not need as an input a lower bound on the minimal expected reward of an arm, and its performance does not scale inversely proportionally to the minimal expected reward. We prove a theoretical regret upper bound to justify these claims. We complement our theoretical results with numerical experiments, showing that the proposed algorithm outperforms state-of-the-art in practice as well.