English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Offline reinforcement learning (ORL) has gained attention as a means oftraining reinforcement learning models using pre-collected static data. Toaddress the issue of limited data and improve downstream ORL performance,recent work has attempted to expand the dataset's coverage through dataaugmentation. However, most of these methods are tied to a specific policy(policy-dependent), where the generated data can only guarantee to support thecurrent downstream ORL policy, limiting its usage scope on other downstreampolicies. Moreover, the quality of synthetic data is often not well-controlled,which limits the potential for further improving the downstream policy. Totackle these issues, we propose \textbf{HI}gh-quality\textbf{PO}licy-\textbf{DE}coupled~(HIPODE), a novel data augmentation methodfor ORL. On the one hand, HIPODE generates high-quality synthetic data byselecting states near the dataset distribution with potentially high valueamong candidate states using the negative sampling technique. On the otherhand, HIPODE is policy-decoupled, thus can be used as a common plug-in methodfor any downstream ORL process. We conduct experiments on the widely studiedTD3BC and CQL algorithms, and the results show that HIPODE outperforms thestate-of-the-art policy-decoupled data augmentation method and most prevalentmodel-based ORL methods on D4RL benchmarks.

HIPODE: Enhancing Offline Reinforcement Learning with High-Quality Synthetic Data from a Policy-Decoupled Approach