DiffStitch: Boosting Offline Reinforcement Learning with Diffusion-based Trajectory Stitching

In offline reinforcement learning (RL), the performance of the learned policy highly depends on the quality of offline datasets. However, in many cases, the offline dataset contains very limited optimal trajectories, which poses a challenge for offline RL algorithms as agents must acquire the abilit...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2024-02
Main Authors: Li, Guanghe, Shan, Yixiang, Zhu, Zhengbang, Long, Ting, Zhang, Weinan
Format: Article
Language:English
Subjects:
Algorithms
Data augmentation
Datasets
Machine learning
Stitching
Trajectory optimization
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In offline reinforcement learning (RL), the performance of the learned policy highly depends on the quality of offline datasets. However, in many cases, the offline dataset contains very limited optimal trajectories, which poses a challenge for offline RL algorithms as agents must acquire the ability to transit to high-reward regions. To address this issue, we introduce Diffusion-based Trajectory Stitching (DiffStitch), a novel diffusion-based data augmentation pipeline that systematically generates stitching transitions between trajectories. DiffStitch effectively connects low-reward trajectories with high-reward trajectories, forming globally optimal trajectories to address the challenges faced by offline RL algorithms. Empirical experiments conducted on D4RL datasets demonstrate the effectiveness of DiffStitch across RL methodologies. Notably, DiffStitch demonstrates substantial enhancements in the performance of one-step methods (IQL), imitation learning methods (TD3+BC), and trajectory optimization methods (DT).
ISSN:2331-8422