English

Uncertainty-driven Trajectory Truncation for Data Augmentation in Offline Reinforcement Learning

Machine Learning 2023-07-27 v2 Artificial Intelligence

Abstract

Equipped with the trained environmental dynamics, model-based offline reinforcement learning (RL) algorithms can often successfully learn good policies from fixed-sized datasets, even some datasets with poor quality. Unfortunately, however, it can not be guaranteed that the generated samples from the trained dynamics model are reliable (e.g., some synthetic samples may lie outside of the support region of the static dataset). To address this issue, we propose Trajectory Truncation with Uncertainty (TATU), which adaptively truncates the synthetic trajectory if the accumulated uncertainty along the trajectory is too large. We theoretically show the performance bound of TATU to justify its benefits. To empirically show the advantages of TATU, we first combine it with two classical model-based offline RL algorithms, MOPO and COMBO. Furthermore, we integrate TATU with several off-the-shelf model-free offline RL algorithms, e.g., BCQ. Experimental results on the D4RL benchmark show that TATU significantly improves their performance, often by a large margin. Code is available here.

Keywords

Cite

@article{arxiv.2304.04660,
  title  = {Uncertainty-driven Trajectory Truncation for Data Augmentation in Offline Reinforcement Learning},
  author = {Junjie Zhang and Jiafei Lyu and Xiaoteng Ma and Jiangpeng Yan and Jun Yang and Le Wan and Xiu Li},
  journal= {arXiv preprint arXiv:2304.04660},
  year   = {2023}
}
R2 v1 2026-06-28T09:57:37.913Z