Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2401.08819 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929269724151808 |
|---|---|
| author | Cen, Zhepeng Liu, Zuxin Wang, Zitong Yao, Yihang Lam, Henry Zhao, Ding |
| author_facet | Cen, Zhepeng Liu, Zuxin Wang, Zitong Yao, Yihang Lam, Henry Zhao, Ding |
| contents | Offline reinforcement learning (RL) offers a promising direction for learning policies from pre-collected datasets without requiring further interactions with the environment. However, existing methods struggle to handle out-of-distribution (OOD) extrapolation errors, especially in sparse reward or scarce data settings. In this paper, we propose a novel training algorithm called Conservative Density Estimation (CDE), which addresses this challenge by explicitly imposing constraints on the state-action occupancy stationary distribution. CDE overcomes the limitations of existing approaches, such as the stationary distribution correction method, by addressing the support mismatch issue in marginal importance sampling. Our method achieves state-of-the-art performance on the D4RL benchmark. Notably, CDE consistently outperforms baselines in challenging tasks with sparse rewards or insufficient data, demonstrating the advantages of our approach in addressing the extrapolation error problem in offline RL. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_08819 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Learning from Sparse Offline Datasets via Conservative Density Estimation Cen, Zhepeng Liu, Zuxin Wang, Zitong Yao, Yihang Lam, Henry Zhao, Ding Machine Learning Artificial Intelligence Offline reinforcement learning (RL) offers a promising direction for learning policies from pre-collected datasets without requiring further interactions with the environment. However, existing methods struggle to handle out-of-distribution (OOD) extrapolation errors, especially in sparse reward or scarce data settings. In this paper, we propose a novel training algorithm called Conservative Density Estimation (CDE), which addresses this challenge by explicitly imposing constraints on the state-action occupancy stationary distribution. CDE overcomes the limitations of existing approaches, such as the stationary distribution correction method, by addressing the support mismatch issue in marginal importance sampling. Our method achieves state-of-the-art performance on the D4RL benchmark. Notably, CDE consistently outperforms baselines in challenging tasks with sparse rewards or insufficient data, demonstrating the advantages of our approach in addressing the extrapolation error problem in offline RL. |
| title | Learning from Sparse Offline Datasets via Conservative Density Estimation |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2401.08819 |