{"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00231961","sets":["1164:4619:11539:11571"]},"path":["11571"],"owner":"44499","recid":"231961","title":["複雑な決定境界に対応するためのスタッキングアンサンブル学習器による高速道路SA就業者の感情推定における不均衡データ対策手法の比較"],"pubdate":{"attribute_name":"公開日","attribute_value":"2024-01-18"},"_buckets":{"deposit":"57188843-b05b-4905-ad37-2a409ec45322"},"_deposit":{"id":"231961","pid":{"type":"depid","value":"231961","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"複雑な決定境界に対応するためのスタッキングアンサンブル学習器による高速道路SA就業者の感情推定における不均衡データ対策手法の比較","author_link":["627482","627483","627485","627481","627479","627484","627478","627480"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"複雑な決定境界に対応するためのスタッキングアンサンブル学習器による高速道路SA就業者の感情推定における不均衡データ対策手法の比較"},{"subitem_title":"Comparison of Imbalanced Data Handling Techniques in Emotion Estimation of Expressway Service Area Workers Using Stacking Ensemble Learners for Complex Decision Boundaries","subitem_title_language":"en"}]},"item_type_id":"4","publish_date":"2024-01-18","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"国立研究開発法人産業技術総合研究所 人間拡張研究センター"},{"subitem_text_value":"国立研究開発法人産業技術総合研究所 人間拡張研究センター"},{"subitem_text_value":"国立研究開発法人産業技術総合研究所 人間拡張研究センター"},{"subitem_text_value":"国立研究開発法人産業技術総合研究所 人間拡張研究センター"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"Human Augmentation Research Center, National Institute of Advanced Industrial Science and Technology","subitem_text_language":"en"},{"subitem_text_value":"Human Augmentation Research Center, National Institute of Advanced Industrial Science and Technology","subitem_text_language":"en"},{"subitem_text_value":"Human Augmentation Research Center, National Institute of Advanced Industrial Science and Technology","subitem_text_language":"en"},{"subitem_text_value":"Human Augmentation Research Center, National Institute of Advanced Industrial 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/231961/files/IPSJ-CVIM24236039.pdf","label":"IPSJ-CVIM24236039.pdf"},"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-CVIM24236039.pdf","filesize":[{"value":"1.3 MB"}],"mimetype":"application/pdf","priceinfo":[{"tax":["include_tax"],"price":"0","billingrole":"20"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_login","version_id":"b2dd45a9-9384-4a6e-8308-6991bec27a6e","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2024 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":"佐藤, 章博"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"小木曽, 里樹"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"一刈, 良介"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"蔵田, 武志"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Akihiro, Sato","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Satoki, Ogiso","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Ryosuke, Ichikari","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Takeshi, Kurata","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11131797","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":"2188-8701","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"就業者に対する感情推定は健康経営への活用が期待できるが，業務中の就業者に対して感情推定のためのデータを収集することは困難である．我々は，高速道路サービスエリアの商業施設を対象に商用スマートデバイスを利用した行動計測実験を実施し，脈拍数，身体活動量に加え，業務状況に起因することで感情状態と関係があると考えられる屋内位置情報のデータを収集し，これらを用いて就業者の感情推定に取り組んだ．推定の対象として，就業者の主観情報を絵文字を用いた経験サンプリングにより就業中の任意協力で収集した．そこで正解ラベルの数に限りがある上，クラス間の数量に偏りも生じた．さらに，複数の異なる状況から同一の感情状態を示す主観情報がもたらされたと考えられるため，本研究では，分布が複雑なデータの分離に強みを持つアンサンブル学習に着目し，不均衡データ対策を施したスタッキングアンサンブル学習による感情推定の精度を評価した．その結果，不均衡データ対策にデータ拡張を用いた場合と比較して，復元抽出時にアンダーサンプリングを行うバギングを用いた場合においてより高い精度を示した．","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Estimating emotions of workers is promising to promote health and productivity management, while it has difficulty in collecting data to use estimation during work. We conducted experiments to measure workers’ activities by commercial off-the-shelf smart devices targeting commercial facility in an expressway service area, and addressed emotion estimation of workers’ using multimodal data such as pulse rate and physical activity, along with indoor location information, which can be supposed to relate with the context and situation of work. As targets of estimation, we collected subjective information of workers by experience sampling method using emoji through voluntary cooperation during work. Thus, the amount of collected ground truth was limited, and also there was a bias in the number of labels between the emotional states. Since it can be inferred that the ground truth data of same kind emotion brought from multiple different situations, we focused on ensemble learning, which has an advantage in separating data with complex decision boundaries. In this study we evaluated the accuracy of emotion estimation by stacking-based ensemble learning with countermeasures against imbalanced data. The results showed that the estimation accuracy when using a combination of undersampling and bagging is better than when using data augmentation.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告コンピュータビジョンとイメージメディア（CVIM）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2024-01-18","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"39","bibliographicVolumeNumber":"2024-CVIM-236"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"id":231961,"updated":"2025-01-19T10:35:15.837355+00:00","links":{},"created":"2025-01-19T01:32:32.594946+00:00"}