{"links":{},"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00226520","sets":["1164:2735:11166:11285"]},"path":["11285"],"owner":"44499","recid":"226520","title":["MMアルゴリズムによる行列式点過程の学習"],"pubdate":{"attribute_name":"公開日","attribute_value":"2023-06-22"},"_buckets":{"deposit":"35ce5e1b-1c42-40d3-bdea-4925bf619388"},"_deposit":{"id":"226520","pid":{"type":"depid","value":"226520","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"MMアルゴリズムによる行列式点過程の学習","author_link":["601597","601596","601595","601598"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"MMアルゴリズムによる行列式点過程の学習"},{"subitem_title":"Minorization-Maximization for Determinantal Point Processes","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"IBISML","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2023-06-22","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"総合研究大学院大学複合科学研究科統計科学専攻"},{"subitem_text_value":"統計数理研究所／理研AIP"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Department of Statistical Science, SOKENDAI","subitem_text_language":"en"},{"subitem_text_value":" Institute of Statistical Mathematics / RIKEN AIP","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/226520/files/IPSJ-MPS23143052.pdf","label":"IPSJ-MPS23143052.pdf"},"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-MPS23143052.pdf","filesize":[{"value":"1.8 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"17"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_login","version_id":"8626f076-0da7-4a31-a718-9d2b901f95f8","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2023 by the Institute of Electronics, Information and Communication Engineers This SIG report is only available to those in membership of the SIG."}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"川島, 貴大"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"日野, 英逸"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Takahiro, Kawashima","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Hideitsu, Hino","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10505667","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":"2188-8833","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"行列式点過程はある全体集合から多様な要素をもつ部分集合をランダムに生成する確率モデルであり，一般に正定値カーネルによって特徴づけられる．とくに有限集合上の行列式点過程はカーネル行列によってパラメトライズされることから近年機械学習コミュニティにおいて研究が進められており，応用範囲も広がりつつあるが，その効率的な学習法についてはいまだ研究の余地がある．本研究では MM アルゴリズムに基づいて，行列式点過程の最尤推定問題を解くための単調かつ簡潔な学習則を提案する．また MM アルゴリズムの代理関数について，提案法が既存法より局所的にタイトに目的関数を抑えることを示した．人工データおよび実データに対する実験を通し，多くの実験設定において提案法による学習が既存法よりも高速に収束することを確かめた．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"A determinantal point process (DPP) is a powerful probabilistic model that generates diverse random subsets from a ground set. Since a DPP is characterized by a positive definite kernel, a DPP on a finite ground set can be parameterized by a kernel matrix. Recently, DPPs have gained attention in the machine learning community and have been applied to various practical problems; however, there is still room for further research on the learning of DPPs. In this paper, we propose a simple learning rule for full-rank DPPs based on a minorization-maximization (MM) algorithm, which monotonically increases the likelihood in each iteration. We show that our minorizer of the MM algorithm provides a tighter lower-bound compared to an existing method locally. In our experiments on both synthetic and real-world datasets, our method outperforms existing methods in most settings.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"9","bibliographic_titles":[{"bibliographic_title":"研究報告数理モデル化と問題解決（MPS）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2023-06-22","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"52","bibliographicVolumeNumber":"2023-MPS-143"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T01:25:58.118996+00:00","updated":"2025-01-19T12:27:03.737404+00:00","id":226520}