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 Plus annual

Presents the access of premium content as premium feature

Premium Content

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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In recent years, data-driven reinforcement learning (RL), also known asoffline RL, have gained significant attention. However, the role of datasampling techniques in offline RL has been overlooked despite its potential toenhance online RL performance. Recent research suggests applying samplingtechniques directly to state-transitions does not consistently improveperformance in offline RL. Therefore, in this study, we propose a memorytechnique, (Prioritized) Trajectory Replay (TR/PTR), which extends the samplingperspective to trajectories for more comprehensive information extraction fromlimited data. TR enhances learning efficiency by backward sampling oftrajectories that optimizes the use of subsequent state information. Buildingon TR, we build the weighted critic target to avoid sampling unseen actions inoffline training, and Prioritized Trajectory Replay (PTR) that enables moreefficient trajectory sampling, prioritized by various trajectory prioritymetrics. We demonstrate the benefits of integrating TR and PTR with existingoffline RL algorithms on D4RL. In summary, our research emphasizes thesignificance of trajectory-based data sampling techniques in enhancing theefficiency and performance of offline RL algorithms.

Prioritized Trajectory Replay: A Replay Memory for Data-driven Reinforcement Learning