2022-11-28T13:52:40Zhttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_oaipmhoai:ipsj.ixsq.nii.ac.jp:000912782020-10-27T05:02:56Z00934:00989:07128:07129
SemiCCA: Efficient Semi-supervised Learning of Canonical CorrelationsSemiCCA: Efficient Semi-supervised Learning of Canonical Correlationseng[オリジナル論文] Canonical correlation analysis, semi-supervised learning, generalized eigenproblem, principal component analysis, multi-label predictionhttp://id.nii.ac.jp/1001/00091261/Articlehttps://ipsj.ixsq.nii.ac.jp/ej/?action=repository_action_common_download&item_id=91278&item_no=1&attribute_id=1&file_no=1Copyright (c) 2013 by the Information Processing Society of JapanNTT Communication Science Laboratories, NTT CorporationGraduate School of Information Science and Engineering, Tokyo Institute of TechnologyNTT Communication Science Laboratories, NTT Corporation／Graduate School of Information Science and Technologies, the University of TokyoNTT Communication Science Laboratories, NTT Corporation／Graduate School of Information Science and Technologies, the University of TokyoNTT Communication Science Laboratories, NTT CorporationNTT Communication Science Laboratories, NTT CorporationNTT Communication Science Laboratories, NTT CorporationAkisato, KimuraMasashi, SugiyamaTakuho, NakanoHirokazu, KameokaHitoshi, SakanoEisaku, MaedaKatsuhiko, IshiguroCanonical correlation analysis (CCA) is a powerful tool for analyzing multi-dimensional paired data. However, CCA tends to perform poorly when the number of paired samples is limited, which is often the case in practice. To cope with this problem, we propose a semi-supervised variant of CCA named SemiCCA that allows us to incorporate additional unpaired samples for mitigating overfitting. Advantages of the proposed method over previously proposed methods are its computational efficiency and intuitive operationality: it smoothly bridges the generalized eigenvalue problems of CCA and principal component analysis (PCA), and thus its solution can be computed efficiently just by solving a single eigenvalue problem as the original CCA.Canonical correlation analysis (CCA) is a powerful tool for analyzing multi-dimensional paired data. However, CCA tends to perform poorly when the number of paired samples is limited, which is often the case in practice. To cope with this problem, we propose a semi-supervised variant of CCA named SemiCCA that allows us to incorporate additional unpaired samples for mitigating overfitting. Advantages of the proposed method over previously proposed methods are its computational efficiency and intuitive operationality: it smoothly bridges the generalized eigenvalue problems of CCA and principal component analysis (PCA), and thus its solution can be computed efficiently just by solving a single eigenvalue problem as the original CCA.AA11464803情報処理学会論文誌数理モデル化と応用（TOM）611281352013-03-121882-77802013-03-07