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

Offline reinforcement learning (RL) can be used to improve future performanceby leveraging historical data. There exist many different algorithms foroffline RL, and it is well recognized that these algorithms, and theirhyperparameter settings, can lead to decision policies with substantiallydiffering performance. This prompts the need for pipelines that allowpractitioners to systematically perform algorithm-hyperparameter selection fortheir setting. Critically, in most real-world settings, this pipeline must onlyinvolve the use of historical data. Inspired by statistical model selectionmethods for supervised learning, we introduce a task- and method-agnosticpipeline for automatically training, comparing, selecting, and deploying thebest policy when the provided dataset is limited in size. In particular, ourwork highlights the importance of performing multiple data splits to producemore reliable algorithm-hyperparameter selection. While this is a commonapproach in supervised learning, to our knowledge, this has not been discussedin detail in the offline RL setting. We show it can have substantial impactswhen the dataset is small. Compared to alternate approaches, our proposedpipeline outputs higher-performing deployed policies from a broad range ofoffline policy learning algorithms and across various simulation domains inhealthcare, education, and robotics. This work contributes toward thedevelopment of a general-purpose meta-algorithm for automaticalgorithm-hyperparameter selection for offline RL.

Data-Efficient Pipeline for Offline Reinforcement Learning with Limited Data