DPO: Differential reinforcement learning with application to optimal configuration search

arXiv:2404.15617v1 Announce Type: new
Abstract: Reinforcement learning (RL) with continuous state and action spaces remains one of the most challenging problems within the field. Most current learning methods focus on integral identities such as value functions to derive an optimal strategy for the learning agent. In this paper, we instead study the dual form of the original RL formulation to propose the first differential RL framework that can handle settings with limited training samples and short-length episodes. Our approach introduces …

abstract agent application arxiv continuous cs.ai cs.lg current differential dpo focus functions integral math.oc math.st paper reinforcement reinforcement learning search spaces state stat.th strategy type value