A Survey of Meta-Reinforcement Learning

While deep reinforcement learning (RL) has fueled multiple high-profile successes in machine learning, it is held back from more widespread adoption by its often poor data efficiency and the limited generality of the policies it produces. A promising approach for alleviating these limitations is to...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Beck, Jacob, Vuorio, Risto, Liu, Evan Zheran, Xiong, Zheng, Zintgraf, Luisa, Finn, Chelsea, Whiteson, Shimon
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Beck, Jacob
Vuorio, Risto
Liu, Evan Zheran
Xiong, Zheng
Zintgraf, Luisa
Finn, Chelsea
Whiteson, Shimon
description While deep reinforcement learning (RL) has fueled multiple high-profile successes in machine learning, it is held back from more widespread adoption by its often poor data efficiency and the limited generality of the policies it produces. A promising approach for alleviating these limitations is to cast the development of better RL algorithms as a machine learning problem itself in a process called meta-RL. Meta-RL is most commonly studied in a problem setting where, given a distribution of tasks, the goal is to learn a policy that is capable of adapting to any new task from the task distribution with as little data as possible. In this survey, we describe the meta-RL problem setting in detail as well as its major variations. We discuss how, at a high level, meta-RL research can be clustered based on the presence of a task distribution and the learning budget available for each individual task. Using these clusters, we then survey meta-RL algorithms and applications. We conclude by presenting the open problems on the path to making meta-RL part of the standard toolbox for a deep RL practitioner.
doi_str_mv 10.48550/arxiv.2301.08028
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2301_08028</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2301_08028</sourcerecordid><originalsourceid>FETCH-LOGICAL-a1158-e4c89cc9b8b2527153834383b910a34862621a628633278057d211cbe3cbf21a3</originalsourceid><addsrcrecordid>eNotzkuLwkAQBOC5eFjUH-DJ3DwlTndnks5RZN0VIoKPc5gZOxLQKLMq679fH3so6lBQfEoNQCcpG6PHNvw2twRJQ6JZI3-o0SRaX8NN7tGpjhZysfFKmrY-BS9HaS9RKTa0TbvvqU5tDz_S_--u2s4-N9PvuFx-zaeTMrYAhmNJPRfeF44dGszBEFP6iCtAW0o5wwzBZsgZEeasTb5DAO-EvKsfC3XV8P37olbn0BxtuFdPcvUi0x8RNzjN</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>A Survey of Meta-Reinforcement Learning</title><source>arXiv.org</source><creator>Beck, Jacob ; Vuorio, Risto ; Liu, Evan Zheran ; Xiong, Zheng ; Zintgraf, Luisa ; Finn, Chelsea ; Whiteson, Shimon</creator><creatorcontrib>Beck, Jacob ; Vuorio, Risto ; Liu, Evan Zheran ; Xiong, Zheng ; Zintgraf, Luisa ; Finn, Chelsea ; Whiteson, Shimon</creatorcontrib><description>While deep reinforcement learning (RL) has fueled multiple high-profile successes in machine learning, it is held back from more widespread adoption by its often poor data efficiency and the limited generality of the policies it produces. A promising approach for alleviating these limitations is to cast the development of better RL algorithms as a machine learning problem itself in a process called meta-RL. Meta-RL is most commonly studied in a problem setting where, given a distribution of tasks, the goal is to learn a policy that is capable of adapting to any new task from the task distribution with as little data as possible. In this survey, we describe the meta-RL problem setting in detail as well as its major variations. We discuss how, at a high level, meta-RL research can be clustered based on the presence of a task distribution and the learning budget available for each individual task. Using these clusters, we then survey meta-RL algorithms and applications. We conclude by presenting the open problems on the path to making meta-RL part of the standard toolbox for a deep RL practitioner.</description><identifier>DOI: 10.48550/arxiv.2301.08028</identifier><language>eng</language><subject>Computer Science - Learning</subject><creationdate>2023-01</creationdate><rights>http://creativecommons.org/licenses/by/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-a1158-e4c89cc9b8b2527153834383b910a34862621a628633278057d211cbe3cbf21a3</citedby></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,780,885</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2301.08028$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2301.08028$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Beck, Jacob</creatorcontrib><creatorcontrib>Vuorio, Risto</creatorcontrib><creatorcontrib>Liu, Evan Zheran</creatorcontrib><creatorcontrib>Xiong, Zheng</creatorcontrib><creatorcontrib>Zintgraf, Luisa</creatorcontrib><creatorcontrib>Finn, Chelsea</creatorcontrib><creatorcontrib>Whiteson, Shimon</creatorcontrib><title>A Survey of Meta-Reinforcement Learning</title><description>While deep reinforcement learning (RL) has fueled multiple high-profile successes in machine learning, it is held back from more widespread adoption by its often poor data efficiency and the limited generality of the policies it produces. A promising approach for alleviating these limitations is to cast the development of better RL algorithms as a machine learning problem itself in a process called meta-RL. Meta-RL is most commonly studied in a problem setting where, given a distribution of tasks, the goal is to learn a policy that is capable of adapting to any new task from the task distribution with as little data as possible. In this survey, we describe the meta-RL problem setting in detail as well as its major variations. We discuss how, at a high level, meta-RL research can be clustered based on the presence of a task distribution and the learning budget available for each individual task. Using these clusters, we then survey meta-RL algorithms and applications. We conclude by presenting the open problems on the path to making meta-RL part of the standard toolbox for a deep RL practitioner.</description><subject>Computer Science - Learning</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNotzkuLwkAQBOC5eFjUH-DJ3DwlTndnks5RZN0VIoKPc5gZOxLQKLMq679fH3so6lBQfEoNQCcpG6PHNvw2twRJQ6JZI3-o0SRaX8NN7tGpjhZysfFKmrY-BS9HaS9RKTa0TbvvqU5tDz_S_--u2s4-N9PvuFx-zaeTMrYAhmNJPRfeF44dGszBEFP6iCtAW0o5wwzBZsgZEeasTb5DAO-EvKsfC3XV8P37olbn0BxtuFdPcvUi0x8RNzjN</recordid><startdate>20230119</startdate><enddate>20230119</enddate><creator>Beck, Jacob</creator><creator>Vuorio, Risto</creator><creator>Liu, Evan Zheran</creator><creator>Xiong, Zheng</creator><creator>Zintgraf, Luisa</creator><creator>Finn, Chelsea</creator><creator>Whiteson, Shimon</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20230119</creationdate><title>A Survey of Meta-Reinforcement Learning</title><author>Beck, Jacob ; Vuorio, Risto ; Liu, Evan Zheran ; Xiong, Zheng ; Zintgraf, Luisa ; Finn, Chelsea ; Whiteson, Shimon</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a1158-e4c89cc9b8b2527153834383b910a34862621a628633278057d211cbe3cbf21a3</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Computer Science - Learning</topic><toplevel>online_resources</toplevel><creatorcontrib>Beck, Jacob</creatorcontrib><creatorcontrib>Vuorio, Risto</creatorcontrib><creatorcontrib>Liu, Evan Zheran</creatorcontrib><creatorcontrib>Xiong, Zheng</creatorcontrib><creatorcontrib>Zintgraf, Luisa</creatorcontrib><creatorcontrib>Finn, Chelsea</creatorcontrib><creatorcontrib>Whiteson, Shimon</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Beck, Jacob</au><au>Vuorio, Risto</au><au>Liu, Evan Zheran</au><au>Xiong, Zheng</au><au>Zintgraf, Luisa</au><au>Finn, Chelsea</au><au>Whiteson, Shimon</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>A Survey of Meta-Reinforcement Learning</atitle><date>2023-01-19</date><risdate>2023</risdate><abstract>While deep reinforcement learning (RL) has fueled multiple high-profile successes in machine learning, it is held back from more widespread adoption by its often poor data efficiency and the limited generality of the policies it produces. A promising approach for alleviating these limitations is to cast the development of better RL algorithms as a machine learning problem itself in a process called meta-RL. Meta-RL is most commonly studied in a problem setting where, given a distribution of tasks, the goal is to learn a policy that is capable of adapting to any new task from the task distribution with as little data as possible. In this survey, we describe the meta-RL problem setting in detail as well as its major variations. We discuss how, at a high level, meta-RL research can be clustered based on the presence of a task distribution and the learning budget available for each individual task. Using these clusters, we then survey meta-RL algorithms and applications. We conclude by presenting the open problems on the path to making meta-RL part of the standard toolbox for a deep RL practitioner.</abstract><doi>10.48550/arxiv.2301.08028</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2301.08028
ispartof
issn
language eng
recordid cdi_arxiv_primary_2301_08028
source arXiv.org
subjects Computer Science - Learning
title A Survey of Meta-Reinforcement Learning
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-29T06%3A49%3A55IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=A%20Survey%20of%20Meta-Reinforcement%20Learning&rft.au=Beck,%20Jacob&rft.date=2023-01-19&rft_id=info:doi/10.48550/arxiv.2301.08028&rft_dat=%3Carxiv_GOX%3E2301_08028%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true