{"links":{},"id":196751,"metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00196751","sets":["6504:9795:9801"]},"path":["9801"],"owner":"6748","recid":"196751","title":["意図的代入法における最適代入値の理論的解析"],"pubdate":{"attribute_name":"公開日","attribute_value":"2019-02-28"},"_buckets":{"deposit":"6e4a1d4a-b49e-465a-8b8b-fa37f5e234dd"},"_deposit":{"id":"196751","pid":{"type":"depid","value":"196751","revision_id":0},"owners":[6748],"status":"published","created_by":6748},"item_title":"意図的代入法における最適代入値の理論的解析","author_link":["471190","471189","471191"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"意図的代入法における最適代入値の理論的解析"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"人工知能と認知科学","subitem_subject_scheme":"Other"}]},"item_type_id":"22","publish_date":"2019-02-28","item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"jpn"}]},"item_22_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"阪府大"},{"subitem_text_value":"阪府大"},{"subitem_text_value":"阪府大"}]},"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/196751/files/IPSJ-Z81-2D-03.pdf","label":"IPSJ-Z81-2D-03.pdf"},"date":[{"dateType":"Available","dateValue":"2019-05-28"}],"format":"application/pdf","filename":"IPSJ-Z81-2D-03.pdf","filesize":[{"value":"249.3 kB"}],"mimetype":"application/pdf","accessrole":"open_date","version_id":"9afa14d0-1e58-4034-980d-e6e3a6c0d75e","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2019 by the Information Processing Society of Japan"}]},"item_22_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"福島, 卓弥"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"長谷川, 拓"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"中島, 智晴"}],"nameIdentifiers":[{}]}]},"item_resource_type":{"attribute_name":"資源タイプ","attribute_value_mlt":[{"resourceuri":"http://purl.org/coar/resource_type/c_5794","resourcetype":"conference paper"}]},"item_22_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN00349328","subitem_source_identifier_type":"NCID"}]},"item_22_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"本論文では，予測時にデータが欠損する場合の機械学習モデルの学習法について検討する．学習用データセットには欠損値がなく完全な情報を持っているが，テストデータセットに欠損値が存在するものと仮定する．この仮定の下で，不完全な入力情報に対してロバストな予測が可能なモデルの学習方法を提案する．提案手法では，モデル学習時に特徴量を特定の確率で決められた固定値に置き換える．理論的に誤差の期待値が最小になる代入値を学習用データセットから推定し，その推定代入値を用いてモデルを学習させる．数値実験では推定代入値を用いたモデルと理論値を代入したモデルやランダムに代入したモデルの近似精度を比較し，有効性を調査する．","subitem_description_type":"Other"}]},"item_22_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"50","bibliographic_titles":[{"bibliographic_title":"第81回全国大会講演論文集"}],"bibliographicPageStart":"49","bibliographicIssueDates":{"bibliographicIssueDate":"2019-02-28","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"1","bibliographicVolumeNumber":"2019"}]},"relation_version_is_last":true,"weko_creator_id":"6748"},"created":"2025-01-19T01:01:15.459748+00:00","updated":"2025-01-19T22:37:20.044340+00:00"}