All learning is Local: Multi-agent Learning in Global Reward Games. Chang, Y., Ho, T., & Kaelbling, L. P abstract bibtex In large multiagent games, partial observability, coordination, and credit assignment persistently plague attempts to design good learning algorithms. We provide a simple and efficient algorithm that in part uses a linear system to model the world from a single agent’s limited perspective, and takes advantage of Kalman filtering to allow an agent to construct a good training signal and learn an effective policy.
@article{chang_all_nodate,
title = {All learning is {Local}: {Multi}-agent {Learning} in {Global} {Reward} {Games}},
abstract = {In large multiagent games, partial observability, coordination, and credit assignment persistently plague attempts to design good learning algorithms. We provide a simple and efficient algorithm that in part uses a linear system to model the world from a single agent’s limited perspective, and takes advantage of Kalman filtering to allow an agent to construct a good training signal and learn an effective policy.},
language = {en},
author = {Chang, Yu-han and Ho, Tracey and Kaelbling, Leslie P},
pages = {8}
}
Downloads: 0
{"_id":"5qAJnGfxAvTvipHJn","bibbaseid":"chang-ho-kaelbling-alllearningislocalmultiagentlearninginglobalrewardgames","authorIDs":[],"author_short":["Chang, Y.","Ho, T.","Kaelbling, L. P"],"bibdata":{"bibtype":"article","type":"article","title":"All learning is Local: Multi-agent Learning in Global Reward Games","abstract":"In large multiagent games, partial observability, coordination, and credit assignment persistently plague attempts to design good learning algorithms. We provide a simple and efficient algorithm that in part uses a linear system to model the world from a single agent’s limited perspective, and takes advantage of Kalman filtering to allow an agent to construct a good training signal and learn an effective policy.","language":"en","author":[{"propositions":[],"lastnames":["Chang"],"firstnames":["Yu-han"],"suffixes":[]},{"propositions":[],"lastnames":["Ho"],"firstnames":["Tracey"],"suffixes":[]},{"propositions":[],"lastnames":["Kaelbling"],"firstnames":["Leslie","P"],"suffixes":[]}],"pages":"8","bibtex":"@article{chang_all_nodate,\n\ttitle = {All learning is {Local}: {Multi}-agent {Learning} in {Global} {Reward} {Games}},\n\tabstract = {In large multiagent games, partial observability, coordination, and credit assignment persistently plague attempts to design good learning algorithms. We provide a simple and efficient algorithm that in part uses a linear system to model the world from a single agent’s limited perspective, and takes advantage of Kalman filtering to allow an agent to construct a good training signal and learn an effective policy.},\n\tlanguage = {en},\n\tauthor = {Chang, Yu-han and Ho, Tracey and Kaelbling, Leslie P},\n\tpages = {8}\n}\n\n","author_short":["Chang, Y.","Ho, T.","Kaelbling, L. P"],"key":"chang_all_nodate","id":"chang_all_nodate","bibbaseid":"chang-ho-kaelbling-alllearningislocalmultiagentlearninginglobalrewardgames","role":"author","urls":{},"downloads":0,"html":""},"bibtype":"article","biburl":"https://bibbase.org/zotero/asneha213","creationDate":"2019-07-08T00:48:54.933Z","downloads":0,"keywords":[],"search_terms":["learning","local","multi","agent","learning","global","reward","games","chang","ho","kaelbling"],"title":"All learning is Local: Multi-agent Learning in Global Reward Games","year":null,"dataSources":["fjacg9txEnNSDwee6"]}