Hierarchical Reinforcement Learning with Hindsight

Andrew Levy; Robert Platt; Kate Saenko

Hierarchical Reinforcement Learning with Hindsight

Machine Learning 2019-03-11 v2 Artificial Intelligence Neural and Evolutionary Computing Robotics Machine Learning

Authors: Andrew Levy , Robert Platt , Kate Saenko

Abstract

Reinforcement Learning (RL) algorithms can suffer from poor sample efficiency when rewards are delayed and sparse. We introduce a solution that enables agents to learn temporally extended actions at multiple levels of abstraction in a sample efficient and automated fashion. Our approach combines universal value functions and hindsight learning, allowing agents to learn policies belonging to different time scales in parallel. We show that our method significantly accelerates learning in a variety of discrete and continuous tasks.

Keywords

hierarchical reinforcement learning reinforcement learning reinforcement learning from human feedback

Cite

@article{arxiv.1805.08180,
  title  = {Hierarchical Reinforcement Learning with Hindsight},
  author = {Andrew Levy and Robert Platt and Kate Saenko},
  journal= {arXiv preprint arXiv:1805.08180},
  year   = {2019}
}

Comments

Duplicate. See arXiv:1712.00948 "Learning Multi-Level Hierarchies with Hindsight" for latest version

Hierarchical Reinforcement Learning with Hindsight

Abstract

Keywords

Cite

Comments

Related papers