Matches in SemOpenAlex for { <https://semopenalex.org/work/W4282007775> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4282007775 abstract "In multi-agent reinforcement learning, the use of a global objective is a powerful tool for incentivising cooperation. Unfortunately, it is not sample-efficient to train individual agents with a global reward, because it does not necessarily correlate with an agent's individual actions. This problem can be solved by factorising the global value function into local value functions. Early work in this domain performed factorisation by conditioning local value functions purely on local information. Recently, it has been shown that providing both local information and an encoding of the global state can promote cooperative behaviour. In this paper we propose QGNN, the first value factorisation method to use a graph neural network (GNN) based model. The multi-layer message passing architecture of QGNN provides more representational complexity than models in prior work, allowing it to produce a more effective factorisation. QGNN also introduces a permutation invariant mixer which is able to match the performance of other methods, even with significantly fewer parameters. We evaluate our method against several baselines, including QMIX-Att, GraphMIX, QMIX, VDN, and hybrid architectures. Our experiments include Starcraft, the standard benchmark for credit assignment; Estimate Game, a custom environment that explicitly models inter-agent dependencies; and Coalition Structure Generation, a foundational problem with real-world applications. The results show that QGNN outperforms state-of-the-art value factorisation baselines consistently." @default.
- W4282007775 created "2022-06-13" @default.
- W4282007775 creator A5066624177 @default.
- W4282007775 creator A5068296171 @default.
- W4282007775 date "2022-05-25" @default.
- W4282007775 modified "2023-09-25" @default.
- W4282007775 title "QGNN: Value Function Factorisation with Graph Neural Networks" @default.
- W4282007775 doi "https://doi.org/10.48550/arxiv.2205.13005" @default.
- W4282007775 hasPublicationYear "2022" @default.
- W4282007775 type Work @default.
- W4282007775 citedByCount "0" @default.
- W4282007775 crossrefType "posted-content" @default.
- W4282007775 hasAuthorship W4282007775A5066624177 @default.
- W4282007775 hasAuthorship W4282007775A5068296171 @default.
- W4282007775 hasBestOaLocation W42820077751 @default.
- W4282007775 hasConcept C11413529 @default.
- W4282007775 hasConcept C119857082 @default.
- W4282007775 hasConcept C121332964 @default.
- W4282007775 hasConcept C123657996 @default.
- W4282007775 hasConcept C126255220 @default.
- W4282007775 hasConcept C132525143 @default.
- W4282007775 hasConcept C13280743 @default.
- W4282007775 hasConcept C14036430 @default.
- W4282007775 hasConcept C142362112 @default.
- W4282007775 hasConcept C14646407 @default.
- W4282007775 hasConcept C153349607 @default.
- W4282007775 hasConcept C154945302 @default.
- W4282007775 hasConcept C185798385 @default.
- W4282007775 hasConcept C187834632 @default.
- W4282007775 hasConcept C205649164 @default.
- W4282007775 hasConcept C21308566 @default.
- W4282007775 hasConcept C24890656 @default.
- W4282007775 hasConcept C33923547 @default.
- W4282007775 hasConcept C41008148 @default.
- W4282007775 hasConcept C50644808 @default.
- W4282007775 hasConcept C78458016 @default.
- W4282007775 hasConcept C80444323 @default.
- W4282007775 hasConcept C86803240 @default.
- W4282007775 hasConcept C97541855 @default.
- W4282007775 hasConceptScore W4282007775C11413529 @default.
- W4282007775 hasConceptScore W4282007775C119857082 @default.
- W4282007775 hasConceptScore W4282007775C121332964 @default.
- W4282007775 hasConceptScore W4282007775C123657996 @default.
- W4282007775 hasConceptScore W4282007775C126255220 @default.
- W4282007775 hasConceptScore W4282007775C132525143 @default.
- W4282007775 hasConceptScore W4282007775C13280743 @default.
- W4282007775 hasConceptScore W4282007775C14036430 @default.
- W4282007775 hasConceptScore W4282007775C142362112 @default.
- W4282007775 hasConceptScore W4282007775C14646407 @default.
- W4282007775 hasConceptScore W4282007775C153349607 @default.
- W4282007775 hasConceptScore W4282007775C154945302 @default.
- W4282007775 hasConceptScore W4282007775C185798385 @default.
- W4282007775 hasConceptScore W4282007775C187834632 @default.
- W4282007775 hasConceptScore W4282007775C205649164 @default.
- W4282007775 hasConceptScore W4282007775C21308566 @default.
- W4282007775 hasConceptScore W4282007775C24890656 @default.
- W4282007775 hasConceptScore W4282007775C33923547 @default.
- W4282007775 hasConceptScore W4282007775C41008148 @default.
- W4282007775 hasConceptScore W4282007775C50644808 @default.
- W4282007775 hasConceptScore W4282007775C78458016 @default.
- W4282007775 hasConceptScore W4282007775C80444323 @default.
- W4282007775 hasConceptScore W4282007775C86803240 @default.
- W4282007775 hasConceptScore W4282007775C97541855 @default.
- W4282007775 hasLocation W42820077751 @default.
- W4282007775 hasOpenAccess W4282007775 @default.
- W4282007775 hasPrimaryLocation W42820077751 @default.
- W4282007775 hasRelatedWork W2151702863 @default.
- W4282007775 hasRelatedWork W2729602312 @default.
- W4282007775 hasRelatedWork W2775408020 @default.
- W4282007775 hasRelatedWork W3022038857 @default.
- W4282007775 hasRelatedWork W3040891685 @default.
- W4282007775 hasRelatedWork W3214094365 @default.
- W4282007775 hasRelatedWork W4226345898 @default.
- W4282007775 hasRelatedWork W4303494752 @default.
- W4282007775 hasRelatedWork W4319083788 @default.
- W4282007775 hasRelatedWork W1629725936 @default.
- W4282007775 isParatext "false" @default.
- W4282007775 isRetracted "false" @default.
- W4282007775 workType "article" @default.