{"updated":"2025-01-19T16:39:49.002687+00:00","metadata":{"_oai":{"id":"oai:ipsj.ixsq.nii.ac.jp:00214328","sets":["1164:2836:10501:10733"]},"path":["10733"],"owner":"44499","recid":"214328","title":["Difficulty of detecting overstated dataset size in Federated Learning"],"pubdate":{"attribute_name":"公開日","attribute_value":"2021-12-13"},"_buckets":{"deposit":"182fa24d-5b32-4fe1-b81e-720a97a39388"},"_deposit":{"id":"214328","pid":{"type":"depid","value":"214328","revision_id":0},"owners":[44499],"status":"published","created_by":44499},"item_title":"Difficulty of detecting overstated dataset size in Federated Learning","author_link":["549956","549958","549959","549957","549955","549960"],"item_titles":{"attribute_name":"タイトル","attribute_value_mlt":[{"subitem_title":"Difficulty of detecting overstated dataset size in Federated Learning"},{"subitem_title":"Difficulty of detecting overstated dataset size in Federated Learning","subitem_title_language":"en"}]},"item_keyword":{"attribute_name":"キーワード","attribute_value_mlt":[{"subitem_subject":"機械学習・予測モデル","subitem_subject_scheme":"Other"}]},"item_type_id":"4","publish_date":"2021-12-13","item_4_text_3":{"attribute_name":"著者所属","attribute_value_mlt":[{"subitem_text_value":"The University of Tokyo"},{"subitem_text_value":"Nara Institute of Science and Technology"},{"subitem_text_value":"Nara Institute of Science and Technology"}]},"item_4_text_4":{"attribute_name":"著者所属(英)","attribute_value_mlt":[{"subitem_text_value":"The University of Tokyo","subitem_text_language":"en"},{"subitem_text_value":"Nara Institute of Science and Technology","subitem_text_language":"en"},{"subitem_text_value":"Nara Institute of Science and Technology","subitem_text_language":"en"}]},"item_language":{"attribute_name":"言語","attribute_value_mlt":[{"subitem_language":"eng"}]},"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/214328/files/IPSJ-DPS21189010.pdf","label":"IPSJ-DPS21189010.pdf"},"date":[{"dateType":"Available","dateValue":"2023-12-13"}],"format":"application/pdf","billing":["billing_file"],"filename":"IPSJ-DPS21189010.pdf","filesize":[{"value":"1.1 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":"34"},{"tax":["include_tax"],"price":"0","billingrole":"44"}],"accessrole":"open_date","version_id":"c84306c6-a1b2-41f6-9360-c51919b37a5b","displaytype":"detail","licensetype":"license_note","license_note":"Copyright (c) 2021 by the Information Processing Society of Japan"}]},"item_4_creator_5":{"attribute_name":"著者名","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Hideaki, Takahashi"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kohei, Ichikawa"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Keichi, Takahashi"}],"nameIdentifiers":[{}]}]},"item_4_creator_6":{"attribute_name":"著者名(英)","attribute_type":"creator","attribute_value_mlt":[{"creatorNames":[{"creatorName":"Hideaki, Takahashi","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Kohei, Ichikawa","creatorNameLang":"en"}],"nameIdentifiers":[{}]},{"creatorNames":[{"creatorName":"Keichi, Takahashi","creatorNameLang":"en"}],"nameIdentifiers":[{}]}]},"item_4_source_id_9":{"attribute_name":"書誌レコードID","attribute_value_mlt":[{"subitem_source_identifier":"AN10116224","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":"2188-8906","subitem_source_identifier_type":"ISSN"}]},"item_4_description_7":{"attribute_name":"論文抄録","attribute_value_mlt":[{"subitem_description":"Federated learning is a distributed learning method in which multiple clients cooperate to train a model. Each client sends the gradient of its locally trained model to the server, and the server aggregates the received gradients to build a global model. Since federated learning requires many clients to train a high-performance model, researchers have designed incentive mechanisms that distribute rewards to clients to motivate their participation. While most incentive mechanisms distribute rewards according to the contribution of each client often defined by the number of data, little research has been done on the risk that clients try to claim more rewards by overstating the number of data. This paper proposes three possible methods to exaggerate the size of a local dataset: simple exaggeration of the reported number, modification of the batch size during training, and exaggeration of the dataset by Data Augmentation. Using a variety of models and datasets, we show the inadequacy of current anomaly detection methods in identifying such exaggerations.","subitem_description_type":"Other"}]},"item_4_description_8":{"attribute_name":"論文抄録(英)","attribute_value_mlt":[{"subitem_description":"Federated learning is a distributed learning method in which multiple clients cooperate to train a model. Each client sends the gradient of its locally trained model to the server, and the server aggregates the received gradients to build a global model. Since federated learning requires many clients to train a high-performance model, researchers have designed incentive mechanisms that distribute rewards to clients to motivate their participation. While most incentive mechanisms distribute rewards according to the contribution of each client often defined by the number of data, little research has been done on the risk that clients try to claim more rewards by overstating the number of data. This paper proposes three possible methods to exaggerate the size of a local dataset: simple exaggeration of the reported number, modification of the batch size during training, and exaggeration of the dataset by Data Augmentation. Using a variety of models and datasets, we show the inadequacy of current anomaly detection methods in identifying such exaggerations.","subitem_description_type":"Other"}]},"item_4_biblio_info_10":{"attribute_name":"書誌情報","attribute_value_mlt":[{"bibliographicPageEnd":"6","bibliographic_titles":[{"bibliographic_title":"研究報告マルチメディア通信と分散処理（DPS）"}],"bibliographicPageStart":"1","bibliographicIssueDates":{"bibliographicIssueDate":"2021-12-13","bibliographicIssueDateType":"Issued"},"bibliographicIssueNumber":"10","bibliographicVolumeNumber":"2021-DPS-189"}]},"relation_version_is_last":true,"weko_creator_id":"44499"},"created":"2025-01-19T01:15:09.012247+00:00","id":214328,"links":{}}