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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Experimental verification has been the method of choice for verifying thestability of a multi-agent reinforcement learning (MARL) algorithm as thenumber of agents grows and theoretical analysis becomes prohibitively complex.For cooperative agents, where the ultimate goal is to optimize some globalmetric, the stability is usually verified by observing the evolution of theglobal performance metric over time. If the global metric improves andeventually stabilizes, it is considered a reasonable verification of thesystem's stability.  The main contribution of this note is establishing the need for betterexperimental frameworks and measures to assess the stability of large-scaleadaptive cooperative systems. We show an experimental case study where thestability of the global performance metric can be rather deceiving, hiding anunderlying instability in the system that later leads to a significant drop inperformance. We then propose an alternative metric that relies on agents' localpolicies and show, experimentally, that our proposed metric is more effective(than the traditional global performance metric) in exposing the instability ofMARL algorithms.

Why Global Performance is a Poor Metric for Verifying Convergence of Multi-agent Learning