{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00218647","sets":["1164:5352:10882:10963"]},"path":["10963"],"owner":"44499","recid":"218647","title":["表現空間における分類容易性の評価に基づく継続学習分析手法の提案"],"pubdate":{"attribute_name":"公開日","attribute_value":"2022-06-20"},"_buckets":{"deposit":"71755403-327b-4da3-bfa8-d1021418753e"},"_deposit":{"id":"218647","pid":{"type":"depid","value":"218647","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"表現空間における分類容易性の評価に基づく継続学習分析手法の提案","author_link":["569115","569114","569116","569113","569117","569112"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"表現空間における分類容易性の評価に基づく継続学習分析手法の提案"},{"subitem_title":"A Novel Analytical Method for Continual Learning Based on the Ease of Classification in Representation Spaces","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"2022-06-20","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"青山学院大学大学院理工学研究科"},{"subitem_text_value":" 青山学院大学理工学部"},{"subitem_text_value":" 青山学院大学理工学部"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":" Graduate School of Science and Engineering, Aoyama Gakuin University","subitem_text_language":"en"},{"subitem_text_value":"College of Science and Engineering, Aoyama Gakuin University","subitem_text_language":"en"},{"subitem_text_value":"College of Science and Engineering, Aoyama Gakuin University","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/218647/files/IPSJ-BIO22070017.pdf","label":"IPSJ-BIO22070017.pdf"},"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-BIO22070017.pdf","filesize":[{"value":"1.8 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"41"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_login","version_id":"877f1e41-2a84-4caf-b1da-77466c6d8407","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2022 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":"村田, 健悟"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"伊東, 聖矢"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"大原, 剛三"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Kengo, Murata","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Seiya, Ito","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kouzou, Ohara","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA12055912","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":"2188-8590","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"近年，実世界での画像分類モデルの運用のため，時間の経過とともに分類対象クラスが増加する継続学習シナリオについて，深層学習モデルを対象としたさまざまな学習手法が提案されている．これらの手法は，破滅的忘却と呼ばれる，新規学習クラスに対する最適化により生じる既学習クラスに対する識別能力が完全に消失する現象を緩和する効果を持つことが知られている．しかし，忘却の緩和がどのようなメカニズムで達成されているかについては，依然として不明瞭な点が多い．本研究では，継続学習手法の学習特性を明らかにするため，ネットワークの表現空間における分類容易性の評価に基づく分析手法を提案する．本稿では，2 種類の代表的な継続学習手法を対象とした分析を通し，提案分析手法の有効性を示す．","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"8","bibliographic_titles":[{"bibliographic_title":"研究報告バイオ情報学（BIO）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2022-06-20","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"17","bibliographicVolumeNumber":"2022-BIO-70"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":218647,"updated":"2025-01-19T15:05:45.638974+00:00","links":{},"created":"2025-01-19T01:19:00.109589+00:00"}