{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,22]],"date-time":"2026-09-22T07:14:12Z","timestamp":1790061252314,"version":"4.0.1"},"reference-count":59,"publisher":"Oxford University Press (OUP)","issue":"1","license":[{"start":{"date-parts":[[2024,9,2]],"date-time":"2024-09-02T00:00:00Z","timestamp":1725235200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/academic.oup.com\/pages\/standard-publication-reuse-rights"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,2,10]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>This paper introduces a novel framework called Mode-wise Principal Subspace Pursuit (MOP-UP) to extract hidden variations in both the row and column dimensions for matrix data. To enhance the understanding of the framework, we introduce a class of matrix-variate spiked covariance models that serve as inspiration for the development of the MOP-UP algorithm. The MOP-UP algorithm consists of two steps: Average Subspace Capture (ASC) and Alternating Projection. These steps are specifically designed to capture the row-wise and column-wise dimension-reduced subspaces which contain the most informative features of the data. ASC utilizes a novel average projection operator as initialization and achieves exact recovery in the noiseless setting. We analyse the convergence and non-asymptotic error bounds of MOP-UP, introducing a blockwise matrix eigenvalue perturbation bound that proves the desired bound, where classic perturbation bounds fail. The effectiveness and practical merits of the proposed framework are demonstrated through experiments on both simulated and real datasets. Lastly, we discuss generalizations of our approach to higher-order data.<\/jats:p>","DOI":"10.1093\/jrsssb\/qkae088","type":"journal-article","created":{"date-parts":[[2024,9,2]],"date-time":"2024-09-02T03:23:13Z","timestamp":1725247393000},"page":"232-255","source":"Crossref","is-referenced-by-count":3,"title":["Mode-wise principal subspace pursuit and matrix spiked covariance model"],"prefix":"10.1093","volume":"87","author":[{"given":"Runshi","family":"Tang","sequence":"first","affiliation":[{"name":"Department of Statistics, University of Wisconsin-Madison , Madison, WI ,","place":["USA"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ming","family":"Yuan","sequence":"additional","affiliation":[{"name":"Department of Statistics, Columbia University , New York, NY ,","place":["USA"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8721-5252","authenticated-orcid":false,"given":"Anru R","family":"Zhang","sequence":"additional","affiliation":[{"name":"Departments of Biostatistics & Bioinformatics and Computer Science, Duke University , Durham, NC ,","place":["USA"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2024,9,2]]},"reference":[{"issue":"4","key":"2025090120012428000_qkae088-B1","doi-asserted-by":"publisher","first-page":"433","DOI":"10.1002\/wics.v2:4","article-title":"Principal component analysis","volume":"2","author":"Abdi","year":"2010","journal-title":"Wiley Interdisciplinary Reviews: Computational Statistics"},{"key":"2025090120012428000_qkae088-B2","first-page":"485","article-title":"Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data","volume":"9","author":"Banerjee","year":"2008","journal-title":"The Journal of Machine Learning Research"},{"issue":"2","key":"2025090120012428000_qkae088-B3","doi-asserted-by":"publisher","first-page":"1144","DOI":"10.1214\/21-AOS2143","article-title":"Statistical inference for principal components of spiked covariance matrices","volume":"50","author":"Bao","year":"2022","journal-title":"The Annals of Statistics"},{"issue":"5","key":"2025090120012428000_qkae088-B4","doi-asserted-by":"publisher","first-page":"2221","DOI":"10.1214\/16-AOS1443","article-title":"Optimal rates of convergence for noisy sparse phase retrieval via thresholded wirtinger flow","volume":"44","author":"Cai","year":"2016","journal-title":"The Annals of Statistics"},{"issue":"6","key":"2025090120012428000_qkae088-B5","doi-asserted-by":"publisher","first-page":"3074","DOI":"10.1214\/13-AOS1178","article-title":"Sparse PCA: Optimal rates and adaptive estimation","volume":"41","author":"Cai","year":"2013","journal-title":"The Annals of Statistics"},{"issue":"1","key":"2025090120012428000_qkae088-B6","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1214\/17-AOS1541","article-title":"Rate-optimal perturbation bounds for singular subspaces with applications to high-dimensional statistics","volume":"46","author":"Cai","year":"2018","journal-title":"The Annals of Statistics"},{"issue":"4","key":"2025090120012428000_qkae088-B7","first-page":"1","article-title":"Xgboost: Extreme gradient boosting","volume":"1","author":"Chen","year":"2015","journal-title":"R Package Version 0.4-2"},{"issue":"1","key":"2025090120012428000_qkae088-B8","first-page":"479","article-title":"Graph estimation for matrix-variate Gaussian data","volume":"29","author":"Chen","year":"2019","journal-title":"Statistica Sinica"},{"issue":"1","key":"2025090120012428000_qkae088-B9","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1137\/0707001","article-title":"The rotation of eigenvectors by a perturbation. III","volume":"7","author":"Davis","year":"1970","journal-title":"SIAM Journal on Numerical Analysis"},{"issue":"1","key":"2025090120012428000_qkae088-B10","doi-asserted-by":"publisher","first-page":"265","DOI":"10.1093\/biomet\/68.1.265","article-title":"Some matrix-variate distribution theory: Notational considerations and a Bayesian application","volume":"68","author":"Dawid","year":"1981","journal-title":"Biometrika"},{"issue":"4","key":"2025090120012428000_qkae088-B11","doi-asserted-by":"publisher","first-page":"1253","DOI":"10.1137\/S0895479896305696","article-title":"A multilinear singular value decomposition","volume":"21","author":"De Lathauwer","year":"2000","journal-title":"SIAM Journal on Matrix Analysis and Applications"},{"issue":"4","key":"2025090120012428000_qkae088-B12","doi-asserted-by":"publisher","first-page":"1324","DOI":"10.1137\/S0895479898346995","article-title":"On the best rank-1 and rank-(r 1, r 2,\u2026, rn) approximation of higher-order tensors","volume":"21","author":"De Lathauwer","year":"2000","journal-title":"SIAM Journal on Matrix Analysis and Applications"},{"issue":"2","key":"2025090120012428000_qkae088-B13","doi-asserted-by":"publisher","first-page":"387","DOI":"10.1111\/rssb.12247","article-title":"Matrix variate regressions and envelope models","volume":"80","author":"Ding","year":"2018","journal-title":"Journal of the Royal Statistical Society Series B: Statistical Methodology"},{"issue":"4","key":"2025090120012428000_qkae088-B14","doi-asserted-by":"publisher","first-page":"1742","DOI":"10.1214\/17-AOS1601","article-title":"Optimal shrinkage of eigenvalues in the spiked covariance model","volume":"46","author":"Donoho","year":"2018","journal-title":"Annals of Statistics"},{"issue":"2","key":"2025090120012428000_qkae088-B15","doi-asserted-by":"publisher","first-page":"105","DOI":"10.1080\/00949659908811970","article-title":"The MLE algorithm for the matrix normal distribution","volume":"64","author":"Dutilleul","year":"1999","journal-title":"Journal of Statistical Computation and Simulation"},{"issue":"3","key":"2025090120012428000_qkae088-B16","doi-asserted-by":"publisher","first-page":"922","DOI":"10.1214\/09-AOAS236","article-title":"Are a set of microarrays independent of each other?","volume":"3","author":"Efron","year":"2009","journal-title":"The Annals of Applied Statistics"},{"issue":"8","key":"2025090120012428000_qkae088-B17","doi-asserted-by":"publisher","first-page":"5040","DOI":"10.1109\/TIT.2014.2323359","article-title":"The optimal hard threshold for singular values is 4\/3","volume":"60","author":"Gavish","year":"2014","journal-title":"IEEE Transactions on Information Theory"},{"key":"2025090120012428000_qkae088-B18","author":"Greenewald","year":"2013"},{"issue":"5","key":"2025090120012428000_qkae088-B19","doi-asserted-by":"publisher","first-page":"901","DOI":"10.1111\/rssb.12339","article-title":"Tensor graphical Lasso (TeraLasso)","volume":"81","author":"Greenewald","year":"2019","journal-title":"Journal of the Royal Statistical Society Series B: Statistical Methodology"},{"issue":"5","key":"2025090120012428000_qkae088-B20","doi-asserted-by":"publisher","first-page":"1666","DOI":"10.1111\/rssb.12547","article-title":"Exact clustering in tensor block model: Statistical optimality and computational limit","volume":"84","author":"Han","year":"2022","journal-title":"Journal of the Royal Statistical Society Series B: Statistical Methodology"},{"issue":"1","key":"2025090120012428000_qkae088-B21","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1214\/21-AOS2061","article-title":"An optimal statistical and computational framework for generalized tensor estimation","volume":"50","author":"Han","year":"2022","journal-title":"The Annals of Statistics"},{"key":"2025090120012428000_qkae088-B22","author":"He","year":"2005"},{"issue":"1\u20134","key":"2025090120012428000_qkae088-B23","doi-asserted-by":"publisher","first-page":"164","DOI":"10.1002\/sapm.v6.1","article-title":"The expression of a tensor or a polyadic as a sum of products","volume":"6","author":"Hitchcock","year":"1927","journal-title":"Journal of Mathematics and Physics"},{"issue":"5","key":"2025090120012428000_qkae088-B24","doi-asserted-by":"publisher","first-page":"1659","DOI":"10.1093\/jrsssb\/qkad070","article-title":"Core shrinkage covariance estimation for matrix-variate data","volume":"85","author":"Hoff","year":"2022","journal-title":"Journal of the Royal Statistical Society Series B: Statistical Methodology"},{"issue":"3","key":"2025090120012428000_qkae088-B25","doi-asserted-by":"publisher","first-page":"1169","DOI":"10.1214\/15-AOAS839","article-title":"Multilinear tensor regression for longitudinal relational data","volume":"9","author":"Hoff","year":"2015","journal-title":"The Annals of Applied Statistics"},{"key":"2025090120012428000_qkae088-B26","author":"Inoue","year":"2009"},{"issue":"6","key":"2025090120012428000_qkae088-B27","doi-asserted-by":"publisher","first-page":"3181","DOI":"10.1214\/21-AOS2079","article-title":"Community detection on mixture multilayer networks via regularized tensor decomposition","volume":"49","author":"Jing","year":"2021","journal-title":"The Annals of Statistics"},{"issue":"2","key":"2025090120012428000_qkae088-B28","doi-asserted-by":"publisher","first-page":"295","DOI":"10.1214\/aos\/1009210544","article-title":"On the distribution of the largest eigenvalue in principal components analysis","volume":"29","author":"Johnstone","year":"2001","journal-title":"Annals of Statistics"},{"key":"2025090120012428000_qkae088-B29","doi-asserted-by":"publisher","DOI":"10.1002\/0470013192.bsa501","volume-title":"Encyclopedia of statistics in behavioral science","author":"Jolliffe","year":"2005"},{"issue":"3","key":"2025090120012428000_qkae088-B30","doi-asserted-by":"publisher","first-page":"455","DOI":"10.1137\/07070111X","article-title":"Tensor decompositions and applications","volume":"51","author":"Kolda","year":"2009","journal-title":"SIAM Review"},{"issue":"1","key":"2025090120012428000_qkae088-B31","doi-asserted-by":"publisher","first-page":"464","DOI":"10.1214\/19-AOS1816","article-title":"Efficient estimation of linear functionals of principal components","volume":"48","author":"Koltchinskii","year":"2020","journal-title":"The Annals of Statistics"},{"issue":"4","key":"2025090120012428000_qkae088-B32","doi-asserted-by":"publisher","first-page":"1976","DOI":"10.1214\/15-AIHP705","article-title":"Asymptotics and concentration bounds for bilinear forms of spectral projectors of sample covariance","volume":"52","author":"Koltchinskii","year":"2016","journal-title":"Annales de l'Institut Henri Poincar\u00e9, Probabilit\u00e9s et Statistiques"},{"issue":"1","key":"2025090120012428000_qkae088-B33","doi-asserted-by":"publisher","first-page":"110","DOI":"10.3150\/15-BEJ730","article-title":"Concentration inequalities and moment bounds for sample covariance operators","volume":"23","author":"Koltchinskii","year":"2017","journal-title":"Bernoulli"},{"issue":"5","key":"2025090120012428000_qkae088-B34","doi-asserted-by":"publisher","first-page":"527","DOI":"10.1016\/j.patrec.2004.09.007","article-title":"2D-LDA: A statistical linear discriminant analysis for image matrix","volume":"26","author":"Li","year":"2005","journal-title":"Pattern Recognition Letters"},{"key":"2025090120012428000_qkae088-B35","doi-asserted-by":"publisher","first-page":"485","DOI":"10.1007\/s12561-021-09331-5","article-title":"Characterizing spatiotemporal transcriptome of the human brain via low-rank tensor decomposition","volume":"14","author":"Liu","year":"2022","journal-title":"Statistics in Biosciences"},{"issue":"1","key":"2025090120012428000_qkae088-B36","doi-asserted-by":"publisher","first-page":"18","DOI":"10.1109\/TNN.2007.901277","article-title":"Mpca: Multilinear principal component analysis of tensor objects","volume":"19","author":"Lu","year":"2008","journal-title":"IEEE Transactions on Neural Networks"},{"issue":"7","key":"2025090120012428000_qkae088-B37","doi-asserted-by":"publisher","first-page":"1540","DOI":"10.1016\/j.patcog.2011.01.004","article-title":"A survey of multilinear subspace learning for tensor data","volume":"44","author":"Lu","year":"2011","journal-title":"Pattern Recognition"},{"issue":"3","key":"2025090120012428000_qkae088-B38","doi-asserted-by":"publisher","first-page":"576","DOI":"10.1109\/TASL.2009.2036813","article-title":"Non-negative multilinear principal component analysis of auditory temporal modulations for music genre classification","volume":"18","author":"Panagakis","year":"2009","journal-title":"IEEE Transactions on Audio, Speech, and Language Processing"},{"issue":"4","key":"2025090120012428000_qkae088-B39","first-page":"1617","article-title":"Asymptotics of sample eigenstruture for a large dimensional spiked covariance model","volume":"17","author":"Paul","year":"2007","journal-title":"Statistica Sinica"},{"issue":"11","key":"2025090120012428000_qkae088-B40","doi-asserted-by":"publisher","first-page":"559","DOI":"10.1080\/14786440109462720","article-title":"LIII. On lines and planes of closest fit to systems of points in space","volume":"2","author":"Pearson","year":"1901","journal-title":"The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science"},{"key":"2025090120012428000_qkae088-B41","author":"Richard","year":"2014"},{"issue":"10","key":"2025090120012428000_qkae088-B42","doi-asserted-by":"publisher","first-page":"1397","DOI":"10.1109\/TCSVT.2008.2002825","article-title":"Bayesian tensor approach for 3-D face modeling","volume":"18","author":"Tao","year":"2008","journal-title":"IEEE Transactions on Circuits and Systems for Video Technology"},{"issue":"3","key":"2025090120012428000_qkae088-B43","doi-asserted-by":"publisher","first-page":"611","DOI":"10.1111\/1467-9868.00196","article-title":"Probabilistic principal component analysis","volume":"61","author":"Tipping","year":"1999","journal-title":"Journal of the Royal Statistical Society: Series B (Statistical Methodology)"},{"issue":"21","key":"2025090120012428000_qkae088-B44","doi-asserted-by":"publisher","first-page":"5347","DOI":"10.1109\/TSP.2013.2279355","article-title":"Covariance estimation in high dimensions via Kronecker product expansions","volume":"61","author":"Tsiligkaridis","year":"2013","journal-title":"IEEE Transactions on Signal Processing"},{"issue":"7","key":"2025090120012428000_qkae088-B45","doi-asserted-by":"publisher","first-page":"1743","DOI":"10.1109\/TSP.2013.2240157","article-title":"On convergence of Kronecker graphical Lasso algorithms","volume":"61","author":"Tsiligkaridis","year":"2013","journal-title":"IEEE Transactions on Signal Processing"},{"key":"2025090120012428000_qkae088-B46","doi-asserted-by":"crossref","DOI":"10.1017\/9781108231596","volume-title":"High-dimensional probability: An introduction with applications in data science","author":"Vershynin","year":"2018"},{"issue":"154","key":"2025090120012428000_qkae088-B47","first-page":"1","article-title":"Learning from binary multiway data: Probabilistic tensor decomposition and its statistical optimality","volume":"21","author":"Wang","year":"2020","journal-title":"The Journal of Machine Learning Research"},{"issue":"1","key":"2025090120012428000_qkae088-B48","doi-asserted-by":"publisher","first-page":"131","DOI":"10.1109\/TPAMI.2004.1261097","article-title":"Two-dimensional PCA: A new approach to appearance-based face representation and recognition","volume":"26","author":"Yang","year":"2004","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"2025090120012428000_qkae088-B49","author":"Ye","year":"2004"},{"key":"2025090120012428000_qkae088-B50","author":"Ye","year":"2004"},{"key":"2025090120012428000_qkae088-B51","doi-asserted-by":"publisher","first-page":"119","DOI":"10.1016\/j.jmva.2012.01.005","article-title":"Model selection and estimation in the matrix normal graphical model","volume":"107","author":"Yin","year":"2012","journal-title":"Journal of Multivariate Analysis"},{"issue":"11","key":"2025090120012428000_qkae088-B52","doi-asserted-by":"publisher","first-page":"7311","DOI":"10.1109\/TIT.2018.2841377","article-title":"Tensor SVD: Statistical and computational limits","volume":"64","author":"Zhang","year":"2018","journal-title":"IEEE Transactions on Information Theory"},{"issue":"1","key":"2025090120012428000_qkae088-B53","doi-asserted-by":"publisher","first-page":"107","DOI":"10.1162\/neco_a_01623","article-title":"Cocaine use prediction with tensor-based machine learning on multimodal mri connectome data","volume":"36","author":"Zhang","year":"2023","journal-title":"Neural Computation"},{"issue":"1","key":"2025090120012428000_qkae088-B54","doi-asserted-by":"publisher","first-page":"53","DOI":"10.1214\/21-AOS2074","article-title":"Heteroskedastic PCA: Algorithm, optimality, and applications","volume":"50","author":"Zhang","year":"2022","journal-title":"The Annals of Statistics"},{"key":"2025090120012428000_qkae088-B55","doi-asserted-by":"publisher","DOI":"10.1016\/j.ultramic.2020.113123","article-title":"Denoising atomic resolution 4D scanning transmission electron microscopy data with tensor singular value decomposition","volume":"219","author":"Zhang","year":"2020","journal-title":"Ultramicroscopy"},{"issue":"543","key":"2025090120012428000_qkae088-B56","doi-asserted-by":"publisher","first-page":"1620","DOI":"10.1080\/01621459.2021.2002157","article-title":"Community detection in general hypergraph via graph embedding","volume":"118","author":"Zhen","year":"2022","journal-title":"Journal of the American Statistical Association"},{"issue":"502","key":"2025090120012428000_qkae088-B57","doi-asserted-by":"publisher","first-page":"540","DOI":"10.1080\/01621459.2013.776499","article-title":"Tensor regression with applications in neuroimaging data analysis","volume":"108","author":"Zhou","year":"2013","journal-title":"Journal of the American Statistical Association"},{"issue":"2","key":"2025090120012428000_qkae088-B58","doi-asserted-by":"publisher","first-page":"532","DOI":"10.1214\/13-AOS1187","article-title":"Gemini: Graph estimation with matrix variate normal instances","volume":"42","author":"Zhou","year":"2014","journal-title":"The Annals of Statistics"},{"issue":"6","key":"2025090120012428000_qkae088-B59","doi-asserted-by":"publisher","first-page":"3991","DOI":"10.1109\/TIT.2022.3152733","article-title":"Optimal high-order tensor SVD via tensor-train orthogonal iteration","volume":"68","author":"Zhou","year":"2022","journal-title":"IEEE Transactions on Information Theory"}],"container-title":["Journal of the Royal Statistical Society Series B: Statistical Methodology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/jrsssb\/article-pdf\/87\/1\/232\/58992097\/qkae088.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/jrsssb\/article-pdf\/87\/1\/232\/58992097\/qkae088.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,2]],"date-time":"2025-09-02T00:01:31Z","timestamp":1756771291000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/jrsssb\/article\/87\/1\/232\/7747328"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,9,2]]},"references-count":59,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2024,9,2]]},"published-print":{"date-parts":[[2025,2,10]]}},"URL":"https:\/\/doi.org\/10.1093\/jrsssb\/qkae088","relation":{},"ISSN":["1369-7412","1467-9868"],"issn-type":[{"value":"1369-7412","type":"print"},{"value":"1467-9868","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,2]]},"published":{"date-parts":[[2024,9,2]]}}}