Leveraging Domain-Unlabeled Data in Offline Reinforcement Learning across Two Domains

arXiv:2404.07465v1 Announce Type: new
Abstract: In this paper, we investigate an offline reinforcement learning (RL) problem where datasets are collected from two domains. In this scenario, having datasets with domain labels facilitates efficient policy training. However, in practice, the task of assigning domain labels can be resource-intensive or infeasible at a large scale, leading to a prevalence of domain-unlabeled data. To formalize this challenge, we introduce a novel offline RL problem setting named Positive-Unlabeled Offline RL (PUORL), which incorporates domain-unlabeled …

abstract arxiv cs.lg data datasets domain domains however labels offline paper policy practice reinforcement reinforcement learning training type