Matches in SemOpenAlex for { <https://semopenalex.org/work/W2912104459> ?p ?o ?g. }
Showing items 1 to 76 of
76
with 100 items per page.
- W2912104459 abstract "In this paper, we propose a new learning technique named message-dropout to improve the performance for multi-agent deep reinforcement learning under two application scenarios: 1) classical multi-agent reinforcement learning with direct message communication among agents and 2) centralized training with decentralized execution. In the first application scenario of multi-agent systems in which direct message communication among agents is allowed, the message-dropout technique drops out the received messages from other agents in a block-wise manner with a certain probability in the training phase and compensates for this effect by multiplying the weights of the dropped-out block units with a correction probability. The applied message-dropout technique effectively handles the increased input dimension in multi-agent reinforcement learning with communication and makes learning robust against communication errors in the execution phase. In the second application scenario of centralized training with decentralized execution, we particularly consider the application of the proposed message-dropout to Multi-Agent Deep Deterministic Policy Gradient (MADDPG), which uses a centralized critic to train a decentralized actor for each agent. We evaluate the proposed message-dropout technique for several games, and numerical results show that the proposed message-dropout technique with proper dropout rate improves the reinforcement learning performance significantly in terms of the training speed and the steady-state performance in the execution phase." @default.
- W2912104459 created "2019-02-21" @default.
- W2912104459 creator A5020240958 @default.
- W2912104459 creator A5081545012 @default.
- W2912104459 creator A5085738436 @default.
- W2912104459 date "2019-02-18" @default.
- W2912104459 modified "2023-09-27" @default.
- W2912104459 title "Message-Dropout: An Efficient Training Method for Multi-Agent Deep Reinforcement Learning" @default.
- W2912104459 cites W1522301498 @default.
- W2912104459 cites W1641379095 @default.
- W2912104459 cites W206679605 @default.
- W2912104459 cites W2095705004 @default.
- W2912104459 cites W2100495367 @default.
- W2912104459 cites W2142505627 @default.
- W2912104459 cites W2155968351 @default.
- W2912104459 cites W2294567968 @default.
- W2912104459 cites W2565610523 @default.
- W2912104459 cites W2602275733 @default.
- W2912104459 cites W2768629321 @default.
- W2912104459 cites W2949201811 @default.
- W2912104459 cites W2962938178 @default.
- W2912104459 cites W2963000099 @default.
- W2912104459 doi "https://doi.org/10.48550/arxiv.1902.06527" @default.
- W2912104459 hasPublicationYear "2019" @default.
- W2912104459 type Work @default.
- W2912104459 sameAs 2912104459 @default.
- W2912104459 citedByCount "1" @default.
- W2912104459 countsByYear W29121044592021 @default.
- W2912104459 crossrefType "posted-content" @default.
- W2912104459 hasAuthorship W2912104459A5020240958 @default.
- W2912104459 hasAuthorship W2912104459A5081545012 @default.
- W2912104459 hasAuthorship W2912104459A5085738436 @default.
- W2912104459 hasBestOaLocation W29121044591 @default.
- W2912104459 hasConcept C119857082 @default.
- W2912104459 hasConcept C120314980 @default.
- W2912104459 hasConcept C121332964 @default.
- W2912104459 hasConcept C153294291 @default.
- W2912104459 hasConcept C154945302 @default.
- W2912104459 hasConcept C2524010 @default.
- W2912104459 hasConcept C2776145597 @default.
- W2912104459 hasConcept C2777210771 @default.
- W2912104459 hasConcept C2777211547 @default.
- W2912104459 hasConcept C33923547 @default.
- W2912104459 hasConcept C41008148 @default.
- W2912104459 hasConcept C854659 @default.
- W2912104459 hasConcept C97541855 @default.
- W2912104459 hasConceptScore W2912104459C119857082 @default.
- W2912104459 hasConceptScore W2912104459C120314980 @default.
- W2912104459 hasConceptScore W2912104459C121332964 @default.
- W2912104459 hasConceptScore W2912104459C153294291 @default.
- W2912104459 hasConceptScore W2912104459C154945302 @default.
- W2912104459 hasConceptScore W2912104459C2524010 @default.
- W2912104459 hasConceptScore W2912104459C2776145597 @default.
- W2912104459 hasConceptScore W2912104459C2777210771 @default.
- W2912104459 hasConceptScore W2912104459C2777211547 @default.
- W2912104459 hasConceptScore W2912104459C33923547 @default.
- W2912104459 hasConceptScore W2912104459C41008148 @default.
- W2912104459 hasConceptScore W2912104459C854659 @default.
- W2912104459 hasConceptScore W2912104459C97541855 @default.
- W2912104459 hasLocation W29121044591 @default.
- W2912104459 hasOpenAccess W2912104459 @default.
- W2912104459 hasPrimaryLocation W29121044591 @default.
- W2912104459 hasRelatedWork W1548320035 @default.
- W2912104459 hasRelatedWork W1561560534 @default.
- W2912104459 hasRelatedWork W192062974 @default.
- W2912104459 hasRelatedWork W2354164008 @default.
- W2912104459 hasRelatedWork W2605524926 @default.
- W2912104459 hasRelatedWork W2912104459 @default.
- W2912104459 hasRelatedWork W2964937097 @default.
- W2912104459 hasRelatedWork W3022038857 @default.
- W2912104459 hasRelatedWork W4210794429 @default.
- W2912104459 hasRelatedWork W4280592718 @default.
- W2912104459 isParatext "false" @default.
- W2912104459 isRetracted "false" @default.
- W2912104459 magId "2912104459" @default.
- W2912104459 workType "article" @default.