WEKO3
アイテム
{"_buckets": {"deposit": "be89fc03-fee3-4185-a09b-9607928cc8db"}, "_deposit": {"created_by": 31, "id": "2393", "owners": [31], "pid": {"revision_id": 0, "type": "depid", "value": "2393"}, "status": "published"}, "_oai": {"id": "oai:oist.repo.nii.ac.jp:00002393", "sets": ["226"]}, "author_link": ["15882", "15881", "15883", "15879", "15878", "15880"], "item_10001_biblio_info_7": {"attribute_name": "Bibliographic Information", "attribute_value_mlt": [{"bibliographicIssueDates": {"bibliographicIssueDate": "2021-07-07", "bibliographicIssueDateType": "Issued"}, "bibliographicPageEnd": "394", "bibliographicPageStart": "383", "bibliographicVolumeNumber": "459", "bibliographic_titles": [{}, {"bibliographic_title": "Neurocomputing", "bibliographic_titleLang": "en"}]}]}, "item_10001_creator_3": {"attribute_name": "Author", "attribute_type": "creator", "attribute_value_mlt": [{"creatorNames": [{"creatorName": "Zhang, Xianjie"}], "nameIdentifiers": [{"nameIdentifier": "15878", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Liu, Yu"}], "nameIdentifiers": [{"nameIdentifier": "15879", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Xu, Xiujuan"}], "nameIdentifiers": [{"nameIdentifier": "15880", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Huang, Qiong"}], "nameIdentifiers": [{"nameIdentifier": "15881", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Mao, Hangyu"}], "nameIdentifiers": [{"nameIdentifier": "15882", "nameIdentifierScheme": "WEKO"}]}, {"creatorNames": [{"creatorName": "Carie, Anil"}], "nameIdentifiers": [{"nameIdentifier": "15883", "nameIdentifierScheme": "WEKO"}]}]}, "item_10001_description_5": {"attribute_name": "Abstract", "attribute_value_mlt": [{"subitem_description": "Multi-agent reinforcement learning (MARL) is essential for a wide range of high-dimensional scenarios and complicated tasks with multiple agents. Many attempts have been made for agents with prior domain knowledge and predefined structure. However, the interaction relationship between agents in a multi-agent system (MAS) in general is usually unknown, and previous methods could not tackle dynamical activities in an ever-changing environment. Here we propose a multi-agent Actor-Critic algorithm called Structural Relational Inference Actor-Critic (SRI-AC), which is based on the framework of centralized training and decentralized execution. SRI-AC utilizes the latent codes in variational autoencoder (VAE) to represent interactions between paired agents, and the reconstruction error is based on Graph Neural Network (GNN). With this framework, we test whether the reinforcement learning learners could form an interpretable structure while achieving better performance in both cooperative and competitive scenarios. The results indicate that SRI-AC could be applied to complex dynamic environments to find an interpretable structure while obtaining better performance compared to baseline algorithms.", "subitem_description_type": "Other"}]}, "item_10001_publisher_8": {"attribute_name": "Publisher", "attribute_value_mlt": [{"subitem_publisher": "Elsevier"}]}, "item_10001_relation_14": {"attribute_name": "DOI", "attribute_value_mlt": [{"subitem_relation_type": "isVersionOf", "subitem_relation_type_id": {"subitem_relation_type_id_text": "info:doi/10.1016/j.neucom.2021.07.014", "subitem_relation_type_select": "DOI"}}]}, "item_10001_relation_17": {"attribute_name": "Related site", "attribute_value_mlt": [{"subitem_relation_type_id": {"subitem_relation_type_id_text": "https://www.sciencedirect.com/science/article/pii/S0925231221010481?via%3Dihub", "subitem_relation_type_select": "URI"}}]}, "item_10001_rights_15": {"attribute_name": "Rights", "attribute_value_mlt": [{"subitem_rights": "This article/chapter was published in Neurocomputing, 459, X. J. Zhang, Y. Liu, X. J. Xu, Q. Huang, H. Y. Mao and A. Carie, Structural relational inference actor-critic for multi-agent reinforcement learning, 383-394, Copyright Elsevier 2021."}]}, "item_10001_source_id_9": {"attribute_name": "ISSN", "attribute_value_mlt": [{"subitem_source_identifier": "0925-2312", "subitem_source_identifier_type": "ISSN"}]}, "item_10001_version_type_20": {"attribute_name": "Author\u0027s flag", "attribute_value_mlt": [{"subitem_version_resource": "http://purl.org/coar/version/c_ab4af688f83e57aa", "subitem_version_type": "AM"}]}, "item_files": {"attribute_name": "ファイル情報", "attribute_type": "file", "attribute_value_mlt": [{"accessrole": "open_date", "date": [{"dateType": "Available", "dateValue": "2023-07-07"}], "displaytype": "detail", "download_preview_message": "", "file_order": 0, "filename": "NEUCOM-D-20-03673_R1.pdf", "filesize": [{"value": "1.3 MB"}], "format": "application/pdf", "future_date_message": "", "is_thumbnail": false, "licensefree": "CC BY-NC-ND 4.0\nCreative Commons Attribution-NonCommercial-NoDerivatives 4.0 International(https://creativecommons.org/licenses/by-nc-nd/4.0/)", "licensetype": "license_free", "mimetype": "application/pdf", "size": 1300000.0, "url": {"label": "NEUCOM-D-20-03673_R1", "url": "https://oist.repo.nii.ac.jp/record/2393/files/NEUCOM-D-20-03673_R1.pdf"}, "version_id": "03a10235-98bd-48ea-af74-7b56bde616f0"}]}, "item_keyword": {"attribute_name": "キーワード", "attribute_value_mlt": [{"subitem_subject": "Multi-agent systems", "subitem_subject_scheme": "Other"}, {"subitem_subject": "Deep reinforcement learning", "subitem_subject_scheme": "Other"}, {"subitem_subject": "Variational autoencoder", "subitem_subject_scheme": "Other"}, {"subitem_subject": "Actor-critic", "subitem_subject_scheme": "Other"}, {"subitem_subject": "Graph neural network", "subitem_subject_scheme": "Other"}]}, "item_language": {"attribute_name": "言語", "attribute_value_mlt": [{"subitem_language": "eng"}]}, "item_resource_type": {"attribute_name": "資源タイプ", "attribute_value_mlt": [{"resourcetype": "journal article", "resourceuri": "http://purl.org/coar/resource_type/c_6501"}]}, "item_title": "Structural relational inference actor-critic for multi-agent reinforcement learning", "item_titles": {"attribute_name": "タイトル", "attribute_value_mlt": [{"subitem_title": "Structural relational inference actor-critic for multi-agent reinforcement learning", "subitem_title_language": "en"}]}, "item_type_id": "10001", "owner": "31", "path": ["226"], "permalink_uri": "https://oist.repo.nii.ac.jp/records/2393", "pubdate": {"attribute_name": "公開日", "attribute_value": "2021-11-29"}, "publish_date": "2021-11-29", "publish_status": "0", "recid": "2393", "relation": {}, "relation_version_is_last": true, "title": ["Structural relational inference actor-critic for multi-agent reinforcement learning"], "weko_shared_id": 31}
Structural relational inference actor-critic for multi-agent reinforcement learning
https://oist.repo.nii.ac.jp/records/2393
https://oist.repo.nii.ac.jp/records/2393933de4a8-1d1f-44a8-8369-b092e0de3dcc
名前 / ファイル | ライセンス | アクション |
---|---|---|
NEUCOM-D-20-03673_R1 (1.3 MB)
|
CC BY-NC-ND 4.0
Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International(https://creativecommons.org/licenses/by-nc-nd/4.0/) |
Item type | 学術雑誌論文 / Journal Article(1) | |||||
---|---|---|---|---|---|---|
公開日 | 2021-11-29 | |||||
タイトル | ||||||
言語 | en | |||||
タイトル | Structural relational inference actor-critic for multi-agent reinforcement learning | |||||
言語 | ||||||
言語 | eng | |||||
キーワード | ||||||
主題Scheme | Other | |||||
主題 | Multi-agent systems | |||||
キーワード | ||||||
主題Scheme | Other | |||||
主題 | Deep reinforcement learning | |||||
キーワード | ||||||
主題Scheme | Other | |||||
主題 | Variational autoencoder | |||||
キーワード | ||||||
主題Scheme | Other | |||||
主題 | Actor-critic | |||||
キーワード | ||||||
主題Scheme | Other | |||||
主題 | Graph neural network | |||||
資源タイプ | ||||||
資源タイプ識別子 | http://purl.org/coar/resource_type/c_6501 | |||||
資源タイプ | journal article | |||||
著者(英) |
Zhang, Xianjie
× Zhang, Xianjie× Liu, Yu× Xu, Xiujuan× Huang, Qiong× Mao, Hangyu× Carie, Anil |
|||||
書誌情報 |
en : Neurocomputing 巻 459, p. 383-394, 発行日 2021-07-07 |
|||||
抄録 | ||||||
内容記述タイプ | Other | |||||
内容記述 | Multi-agent reinforcement learning (MARL) is essential for a wide range of high-dimensional scenarios and complicated tasks with multiple agents. Many attempts have been made for agents with prior domain knowledge and predefined structure. However, the interaction relationship between agents in a multi-agent system (MAS) in general is usually unknown, and previous methods could not tackle dynamical activities in an ever-changing environment. Here we propose a multi-agent Actor-Critic algorithm called Structural Relational Inference Actor-Critic (SRI-AC), which is based on the framework of centralized training and decentralized execution. SRI-AC utilizes the latent codes in variational autoencoder (VAE) to represent interactions between paired agents, and the reconstruction error is based on Graph Neural Network (GNN). With this framework, we test whether the reinforcement learning learners could form an interpretable structure while achieving better performance in both cooperative and competitive scenarios. The results indicate that SRI-AC could be applied to complex dynamic environments to find an interpretable structure while obtaining better performance compared to baseline algorithms. | |||||
出版者 | ||||||
出版者 | Elsevier | |||||
ISSN | ||||||
収録物識別子タイプ | ISSN | |||||
収録物識別子 | 0925-2312 | |||||
DOI | ||||||
関連タイプ | isVersionOf | |||||
識別子タイプ | DOI | |||||
関連識別子 | info:doi/10.1016/j.neucom.2021.07.014 | |||||
権利 | ||||||
権利情報 | This article/chapter was published in Neurocomputing, 459, X. J. Zhang, Y. Liu, X. J. Xu, Q. Huang, H. Y. Mao and A. Carie, Structural relational inference actor-critic for multi-agent reinforcement learning, 383-394, Copyright Elsevier 2021. | |||||
関連サイト | ||||||
識別子タイプ | URI | |||||
関連識別子 | https://www.sciencedirect.com/science/article/pii/S0925231221010481?via%3Dihub | |||||
著者版フラグ | ||||||
出版タイプ | AM | |||||
出版タイプResource | http://purl.org/coar/version/c_ab4af688f83e57aa |