{"id":209853,"created":"2025-01-19T01:11:08.181546+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00209853","sets":["1164:5305:10533:10534"]},"path":["10534"],"owner":"44499","recid":"209853","title":["二人不完全情報ゲーム『ガイスター』におけるCFRおよび利得評価関数を用いた混合戦略AIの研究"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-02-26"},"_buckets":{"deposit":"9e39ff83-a1fa-437d-a364-10bdf26f535e"},"_deposit":{"id":"209853","pid":{"type":"depid","value":"209853","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"二人不完全情報ゲーム『ガイスター』におけるCFRおよび利得評価関数を用いた混合戦略AIの研究","author_link":["530147","530146","530148","530149"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"二人不完全情報ゲーム『ガイスター』におけるCFRおよび利得評価関数を用いた混合戦略AIの研究"},{"subitem_title":"A study on mixed strategy AI using CFR and gain evaluation functions in the two-player imperfect information game ”Geister”","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"ガイスター","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2021-02-26","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"北陸先端科学技術大学院大学"},{"subitem_text_value":"北陸先端科学技術大学院大学"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Japan Advanced Institute of Science and Technology","subitem_text_language":"en"},{"subitem_text_value":"Japan Advanced Institute of Science and Technology","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_publisher":{"attribute_name":"出版者","attribute_value_mlt":[{"subitem_publisher":"情報処理学会","subitem_publisher_language":"ja"}]},"publish_status":"0","weko_shared_id":-1,"item_file_price":{"attribute_name":"Billing file","attribute_type":"file","attribute_value_mlt":[{"url":{"url":"https://ipsj.ixsq.nii.ac.jp/record/209853/files/IPSJ-GI21045004.pdf","label":"IPSJ-GI21045004.pdf"},"date":[{"dateType":"Available","dateValue":"2023-02-26"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-GI21045004.pdf","filesize":[{"value":"1.2 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"660","billingrole":"5"},{"tax":["include_tax"],"price":"330","billingrole":"6"},{"tax":["include_tax"],"price":"0","billingrole":"18"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"a5d6886f-163f-4e9a-9177-f79221154c0f","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"川上, 直人"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"池田, 心"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Naoto, Kawakami","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kokolo, Ikeda","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11362144","subitem_source_identifier_type":"NCID"}]},"item_4_textarea_12":{"attribute_name":"Notice","attribute_value_mlt":[{"subitem_textarea_value":"SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc."}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_18gh","resourcetype":"technical report"}]},"item_4_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"2188-8736","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"近年，囲碁や将棋などの完全情報ゲームだけでなく，麻雀，ポーカーといった不完全情報ゲームにおいて，トッププレイヤの実力を上回るAIが報告されている．Head's up limit porkerでは2015年にナッシュ均衡に近い解が得られた．一方，駒移動型の二人零和不完全情報ゲーム「ガイスター」では，3×2盤・短手数の場合を除き，プログラムが人間の実力を上回るなどの成果が報告されていない．特に，GAT2020杯ガイスターAI大会で優勝した「紫駒AI」は，初期配置以外に確率的な行動が無く，戦略を知っている相手には弱い．例えば，紫駒AIは青駒をめったに相手駒に隣接させず，赤駒をよく隣接させる特徴がある．そのため，隣接させてきた場合には赤駒だろうという推測が容易にできてしまう．そこで本稿では4×4盤ガイスターにおいて，確率的な行動を取らせるためにCFRを用いる．ただし，CFRの計算量を抑えるために数手で木の展開をやめ，未決着なリーフノードにおいて利得評価をおこなう関数を設計する．本稿では，紫駒AIとそれに付け込むAIを含む4種類のベンチマークプレイヤとの対戦実験により，提案手法によるAI2種類の性能評価をおこなった．結果，ベンチマーク同士の対戦では，どのAIについても勝率15%を下回る苦手な相手が存在した一方で，提案手法では最も苦手な対戦相手に対して勝率24%，19%を挙げた．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"7","bibliographic_titles":[{"bibliographic_title":"研究報告ゲーム情報学（GI）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2021-02-26","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"4","bibliographicVolumeNumber":"2021-GI-45"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"updated":"2025-01-19T18:22:09.519972+00:00","links":{}}