English

Imitating Graph-Based Planning with Goal-Conditioned Policies

Machine Learning 2023-03-21 v1 Artificial Intelligence

Abstract

Recently, graph-based planning algorithms have gained much attention to solve goal-conditioned reinforcement learning (RL) tasks: they provide a sequence of subgoals to reach the target-goal, and the agents learn to execute subgoal-conditioned policies. However, the sample-efficiency of such RL schemes still remains a challenge, particularly for long-horizon tasks. To address this issue, we present a simple yet effective self-imitation scheme which distills a subgoal-conditioned policy into the target-goal-conditioned policy. Our intuition here is that to reach a target-goal, an agent should pass through a subgoal, so target-goal- and subgoal- conditioned policies should be similar to each other. We also propose a novel scheme of stochastically skipping executed subgoals in a planned path, which further improves performance. Unlike prior methods that only utilize graph-based planning in an execution phase, our method transfers knowledge from a planner along with a graph into policy learning. We empirically show that our method can significantly boost the sample-efficiency of the existing goal-conditioned RL methods under various long-horizon control tasks.

Keywords

Cite

@article{arxiv.2303.11166,
  title  = {Imitating Graph-Based Planning with Goal-Conditioned Policies},
  author = {Junsu Kim and Younggyo Seo and Sungsoo Ahn and Kyunghwan Son and Jinwoo Shin},
  journal= {arXiv preprint arXiv:2303.11166},
  year   = {2023}
}

Comments

Accepted to ICLR 2023

R2 v1 2026-06-28T09:24:19.852Z