Flooding with Absorption: An Efficient Protocol for Heterogeneous Bandits over Complex Networks

arXiv:2303.05445v4 Announce Type: replace
Abstract: Multi-armed bandits are extensively used to model sequential decision-making, making them ubiquitous in many real-life applications such as online recommender systems and wireless networking. We consider a multi-agent setting where each agent solves their own bandit instance endowed with a different set of arms. Their goal is to minimize their group regret while collaborating via some communication protocol over a given network. Previous literature on this problem only considered arm heterogeneity and networked agents separately. …

abstract agent applications arxiv cs.dc cs.lg cs.ni decision flooding instance life making multi-agent multi-armed bandits networking networks protocol recommender systems set stat.ml systems them type wireless wireless networking