Matches in SemOpenAlex for { <https://semopenalex.org/work/W3089778445> ?p ?o ?g. }
- W3089778445 abstract "Role-based learning holds the promise of achieving scalable multi-agent learning by decomposing complex tasks using roles. However, it is largely unclear how to efficiently discover such a set of roles. To solve this problem, we propose to first decompose joint action spaces into restricted role action spaces by clustering actions according to their effects on the environment and other agents. Learning a role selector based on action effects makes role discovery much easier because it forms a bi-level learning hierarchy -- the role selector searches in a smaller role space and at a lower temporal resolution, while role policies learn in significantly reduced primitive action-observation spaces. We further integrate information about action effects into the role policies to boost learning efficiency and policy generalization. By virtue of these advances, our method (1) outperforms the current state-of-the-art MARL algorithms on 10 of the 14 scenarios that comprise the challenging StarCraft II micromanagement benchmark and (2) achieves rapid transfer to new environments with three times the number of agents. Demonstrative videos are available at this https URL ." @default.
- W3089778445 created "2020-10-08" @default.
- W3089778445 creator A5010176958 @default.
- W3089778445 creator A5012361632 @default.
- W3089778445 creator A5031139059 @default.
- W3089778445 creator A5051483779 @default.
- W3089778445 creator A5056879203 @default.
- W3089778445 creator A5071784141 @default.
- W3089778445 date "2020-10-04" @default.
- W3089778445 modified "2023-09-26" @default.
- W3089778445 title "RODE: Learning Roles to Decompose Multi-Agent Tasks" @default.
- W3089778445 cites W1512681781 @default.
- W3089778445 cites W1540755814 @default.
- W3089778445 cites W154189223 @default.
- W3089778445 cites W1641379095 @default.
- W3089778445 cites W164791823 @default.
- W3089778445 cites W1787758814 @default.
- W3089778445 cites W1992149319 @default.
- W3089778445 cites W1998155433 @default.
- W3089778445 cites W2104602264 @default.
- W3089778445 cites W2109910161 @default.
- W3089778445 cites W2110906765 @default.
- W3089778445 cites W2111877087 @default.
- W3089778445 cites W2111967991 @default.
- W3089778445 cites W2116046643 @default.
- W3089778445 cites W2118318536 @default.
- W3089778445 cites W2130891285 @default.
- W3089778445 cites W2134779831 @default.
- W3089778445 cites W2145805610 @default.
- W3089778445 cites W2160371091 @default.
- W3089778445 cites W2164858444 @default.
- W3089778445 cites W2167670334 @default.
- W3089778445 cites W2177768973 @default.
- W3089778445 cites W2183243664 @default.
- W3089778445 cites W2215378786 @default.
- W3089778445 cites W2404646363 @default.
- W3089778445 cites W2417280899 @default.
- W3089778445 cites W2500451535 @default.
- W3089778445 cites W2556477470 @default.
- W3089778445 cites W2564324149 @default.
- W3089778445 cites W2578961791 @default.
- W3089778445 cites W2617816519 @default.
- W3089778445 cites W2747213132 @default.
- W3089778445 cites W2765172389 @default.
- W3089778445 cites W2768629321 @default.
- W3089778445 cites W2770423710 @default.
- W3089778445 cites W2786572318 @default.
- W3089778445 cites W2789062658 @default.
- W3089778445 cites W2793398421 @default.
- W3089778445 cites W2794643322 @default.
- W3089778445 cites W2803281228 @default.
- W3089778445 cites W2806824292 @default.
- W3089778445 cites W2807741983 @default.
- W3089778445 cites W2885376584 @default.
- W3089778445 cites W2891661335 @default.
- W3089778445 cites W2900926445 @default.
- W3089778445 cites W2901992463 @default.
- W3089778445 cites W2903327785 @default.
- W3089778445 cites W2911743772 @default.
- W3089778445 cites W2913343212 @default.
- W3089778445 cites W2913654295 @default.
- W3089778445 cites W2914351253 @default.
- W3089778445 cites W2920215304 @default.
- W3089778445 cites W2920949972 @default.
- W3089778445 cites W2946606218 @default.
- W3089778445 cites W2947805761 @default.
- W3089778445 cites W2948099544 @default.
- W3089778445 cites W2949267040 @default.
- W3089778445 cites W2951984055 @default.
- W3089778445 cites W2962823158 @default.
- W3089778445 cites W2962938168 @default.
- W3089778445 cites W2962966033 @default.
- W3089778445 cites W2962976158 @default.
- W3089778445 cites W2963000099 @default.
- W3089778445 cites W2963321092 @default.
- W3089778445 cites W2963407617 @default.
- W3089778445 cites W2963588154 @default.
- W3089778445 cites W2963761387 @default.
- W3089778445 cites W2963762747 @default.
- W3089778445 cites W2963875819 @default.
- W3089778445 cites W2964158321 @default.
- W3089778445 cites W2964179661 @default.
- W3089778445 cites W2964332824 @default.
- W3089778445 cites W2964338167 @default.
- W3089778445 cites W2970514967 @default.
- W3089778445 cites W2970534317 @default.
- W3089778445 cites W2971631853 @default.
- W3089778445 cites W2982316857 @default.
- W3089778445 cites W2987204115 @default.
- W3089778445 cites W2994110384 @default.
- W3089778445 cites W2995107092 @default.
- W3089778445 cites W2995520132 @default.
- W3089778445 cites W2996037775 @default.
- W3089778445 cites W2996354442 @default.
- W3089778445 cites W2996549507 @default.
- W3089778445 cites W3010758700 @default.
- W3089778445 cites W3027456239 @default.
- W3089778445 cites W3034971464 @default.
- W3089778445 cites W3035041226 @default.
- W3089778445 cites W3035133122 @default.