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

Peer learning is a novel high-level reinforcement learning framework foragents learning in groups. While standard reinforcement learning trains anindividual agent in trial-and-error fashion, all on its own, peer learningaddresses a related setting in which a group of agents, i.e., peers, learns tomaster a task simultaneously together from scratch. Peers are allowed tocommunicate only about their own states and actions recommended by others:"What would you do in my situation?". Our motivation is to study the learningbehavior of these agents. We formalize the teacher selection process in theaction advice setting as a multi-armed bandit problem and therefore highlightthe need for exploration. Eventually, we analyze the learning behavior of thepeers and observe their ability to rank the agents' performance within thestudy group and understand which agents give reliable advice. Further, wecompare peer learning with single agent learning and a state-of-the-art actionadvice baseline. We show that peer learning is able to outperform single-agentlearning and the baseline in several challenging discrete and continuous OpenAIGym domains. Doing so, we also show that within such a framework complexpolicies from action recommendations beyond discrete action spaces can evolve.

Peer Learning: Learning Complex Policies in Groups from Scratch via Action Recommendations