{"updated":"2025-01-19T23:56:24.839935+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00192959","sets":["934:989:9407:9598"]},"path":["9598"],"owner":"44499","recid":"192959","title":["教師あり学習に基づくGranger causalityの推定"],"pubdate":{"attribute_name":"公開日","attribute_value":"2018-12-20"},"_buckets":{"deposit":"f347b976-2060-430a-a067-10188a4edbf1"},"_deposit":{"id":"192959","pid":{"type":"depid","value":"192959","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"教師あり学習に基づくGranger causalityの推定","author_link":["451654","451653","451652","451655"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"教師あり学習に基づくGranger causalityの推定"},{"subitem_title":"A Supervised Learning Approach to Granger Causality Inference","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"[オリジナル論文] Granger causality，時系列解析，カーネル法","subitem_subject_scheme":"Other"}]},"item_type_id":"3","publish_date":"2018-12-20","item_3_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"NTTコミュニケーション科学基礎研究所"},{"subitem_text_value":"NTTコミュニケーション科学基礎研究所"}]},"item_3_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"NTT Communication Science Laboratories","subitem_text_language":"en"},{"subitem_text_value":"NTT Communication Science Laboratories","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/192959/files/IPSJ-TOM1103007.pdf","label":"IPSJ-TOM1103007.pdf"},"date":[{"dateType":"Available","dateValue":"2020-12-20"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-TOM1103007.pdf","filesize":[{"value":"1.7 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":"17"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"88407947-f49c-4b4c-bff1-d01f83a94aec","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2018 by the Information Processing Society of Japan"}]},"item_3_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"近原, 鷹一"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"藤野, 昭典"}],"nameIdentifiers":[{}]}]},"item_3_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Yoichi, Chikahara","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Akinori, Fujino","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_3_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AA11464803","subitem_source_identifier_type":"NCID"}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_6501","resourcetype":"journal article"}]},"item_3_source_id_11":{"attribute_name":"ISSN","attribute_value_mlt":[{"subitem_source_identifier":"1882-7780","subitem_source_identifier_type":"ISSN"}]},"item_3_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"Granger causalityとは変数間の因果関係の定義の1つであり，その推定は時系列解析における重要なタスクの1つである．従来手法では，回帰モデルを用いてGranger causalityの方向を推定するが，その推定精度は，各々のデータに対して適切な回帰モデルを選択するか否かに強く依存する．しかし，回帰モデルの選択には，データ解析に関する深い専門知識が要求されるため，実際には容易なことではない．本論文では，教師あり学習に基づくGranger causalityの推定手法を提案する．提案手法では，過去の値で条件付けられた条件付き分布間の距離を用いた特徴量表現を用いる．この特徴量表現が，Granger causalityの有無・方向の異なる時系列に対して，十分異なる特徴ベクトルを与えることを，人工データを用いた実験により示す．また，このような特徴ベクトルの差異によって，提案手法が既存手法より高い推定精度を達成したことを，人工データ・実データを用いた実験により示す．","subitem_description_type":"Other"}]},"item_3_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Granger causality is one of the definitions of temporal causality between variables, and inferring Granger causality is an important task in time series analysis. Traditional methods use regression models for this task. Since the inference accuracies of these methods depend largely on whether or not we select an appropriate regression model for each time series data. However, it is not easy because such selection of regression models requires a deep understanding of the data analysis. This paper proposes a supervised learning framework that utilizes a classifier instead of regression models. Our proposed method employs a feature representation that utilizes the distance between the conditional distributions given past variable values. We experimentally show that the feature representation gives sufficiently different feature vectors for time series with different Granger causality. In addition, we confirmed that such difference of feature vectors enables our method to achieve higher inference accuracy than the existing methods.","subitem_description_type":"Other"}]},"item_3_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"73","bibliographic_titles":[{"bibliographic_title":"情報処理学会論文誌数理モデル化と応用（TOM）"}],"bibliographicPageStart":"58","bibliographicIssueDates":{"bibliographicIssueDate":"2018-12-20","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"3","bibliographicVolumeNumber":"11"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T00:58:38.052944+00:00","id":192959,"links":{}}