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

Independent learners are agents that employ single-agent algorithms inmulti-agent systems, intentionally ignoring the effect of other strategicagents. This paper studies mean-field games from a decentralized learningperspective, with two primary objectives: (i) to identify structure that canguide algorithm design, and (ii) to understand the emergent behaviour insystems of independent learners. We study a new model of partially observedmean-field games with finitely many players, local action observability, and ageneral observation channel for partial observations of the global state.Specific observation channels considered include (a) global observability, (b)local and mean-field observability, (c) local and compressed mean-fieldobservability, and (d) only local observability. We establish conditions underwhich the control problem of a given agent is equivalent to a fully observedMDP, as well as conditions under which the control problem is equivalent onlyto a POMDP. Building on the connection to MDPs, we prove the existence ofperfect equilibrium among memoryless stationary policies under mean-fieldobservability. Leveraging the connection to POMDPs, we prove convergence oflearning iterates obtained by independent learning agents under any of theaforementioned observation channels. We interpret the limiting values assubjective value functions, which an agent believes to be relevant to itscontrol problem. These subjective value functions are then used to proposesubjective Q-equilibrium, a new solution concept for partially observedn-player mean-field games, whose existence is proved under mean-field or globalobservability.We provide a decentralized learning algorithm for partiallyobserved n-player mean-field games, and we show that it drives play tosubjective Q-equilibrium by adapting the recently developed theory ofsatisficing paths to allow for subjectivity.

Independent Learning in Mean-Field Games: Satisficing Paths and Convergence to Subjective Equilibria