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

Curriculum Reinforcement Learning (CRL) aims to create a sequence of tasks,starting from easy ones and gradually learning towards difficult tasks. In thiswork, we focus on the idea of framing CRL as interpolations between a source(auxiliary) and a target task distribution. Although existing studies haveshown the great potential of this idea, it remains unclear how to formallyquantify and generate the movement between task distributions. Inspired by theinsights from gradual domain adaptation in semi-supervised learning, we createa natural curriculum by breaking down the potentially large task distributionalshift in CRL into smaller shifts. We propose GRADIENT, which formulates CRL asan optimal transport problem with a tailored distance metric between tasks.Specifically, we generate a sequence of task distributions as a geodesicinterpolation (i.e., Wasserstein barycenter) between the source and targetdistributions. Different from many existing methods, our algorithm considers atask-dependent contextual distance metric and is capable of handlingnonparametric distributions in both continuous and discrete context settings.In addition, we theoretically show that GRADIENT enables smooth transferbetween subsequent stages in the curriculum under certain conditions. Weconduct extensive experiments in locomotion and manipulation tasks and showthat our proposed GRADIENT achieves higher performance than baselines in termsof learning efficiency and asymptotic performance.

Curriculum Reinforcement Learning using Optimal Transport via Gradual Domain Adaptation