Observe Before Play: Multi-armed Bandit with Pre-observations

TL;DR

This paper introduces algorithms for multi-armed bandit problems with pre-observation of rewards, balancing observation costs and reward maximization, and extends to multi-player scenarios with collision management, demonstrating improved regret bounds and practical performance.

Contribution

It proposes the OBP-UCB algorithm for single-player and centralized/distributed algorithms for multi-player bandits with pre-observations, providing theoretical regret bounds and empirical validation.

Findings

OBP-UCB achieves $O(K^2 ext{log} T)$ regret for single-player.

C-MP-OBP attains $O(rac{K^4}{M^2} ext{log} T)$ regret in multi-player setting.

Distributed algorithms outperform heuristics and non-pre-observation policies.

Abstract

We consider the stochastic multi-armed bandit (MAB) problem in a setting where a player can pay to pre-observe arm rewards before playing an arm in each round. Apart from the usual trade-off between exploring new arms to find the best one and exploiting the arm believed to offer the highest reward, we encounter an additional dilemma: pre-observing more arms gives a higher chance to play the best one, but incurs a larger cost. For the single-player setting, we design an Observe-Before-Play Upper Confidence Bound (OBP-UCB) algorithm for $K$ arms with Bernoulli rewards, and prove a $T$-round regret upper bound $O(K^2\log T)$. In the multi-player setting, collisions will occur when players select the same arm to play in the same round. We design a centralized algorithm, C-MP-OBP, and prove its $T$-round regret relative to an offline greedy strategy is upper bounded in $O(\frac{K^4}{M^2}\log T)$ for $K$ arms and $M$ players. We also propose distributed versions of the C-MP-OBP policy, called D-MP-OBP and D-MP-Adapt-OBP, achieving logarithmic regret with respect to collision-free target policies. Experiments on synthetic data and wireless channel traces show that C-MP-OBP and D-MP-OBP outperform random heuristics and offline optimal policies that do not allow pre-observations.