{"type":"ScholarlyArticle","title":"Robust Low-Tubal-Rank Tensor Completion via Convex Optimization","authors":[{"name":"Qiang Jiang","slug":"qiang-jiang","given":"Qiang","family":"Jiang","affiliation":"Department of Mathematics , The University of Hong Kong , Hong Kong; partment of Mathematics , Hong Kong Baptist University."},{"name":"Michael Ng","slug":"michael-ng","given":"Michael","family":"Ng","affiliation":"Department of Mathematics , The University of Hong Kong , Hong Kong"}],"abstract":"This paper considers the problem of recovering multidimensional array, in particular third-order tensor, from a random subset of its arbitrarily corrupted entries. Our study is based on a recently proposed algebraic framework in which the tensor-SVD is introduced to capture the low-tubal-rank structure in tensor. We analyze the performance of a convex program, which minimizes a weighted combination of the tensor nuclear norm, a convex surrogate for the tensor tubal rank, and the tensor l1 norm. We prove that under certain incoherence conditions, this program can recover the tensor exactly with overwhelming probability, provided that its tubal rank is not too large and that the corruptions are reasonably sparse. The number of required observations is order optimal (up to a logarithm factor) when comparing with the degrees of freedom of the low-tubal-rank tensor. Numerical experiments verify our theoretical results and real-world applications demonstrate the effectiveness of our algorithm.","doi":"10.24963/ijcai.2019/368","url":"https://ijcai.org/proceedings/2019/368","pdf":"https://ijcai.org/proceedings/2019/0368.pdf","markdown":"https://ijcai.org/proceedings/2019/368.md","bibtex":"https://ijcai.org/proceedings/2019/bibtex/368","volume":{"slug":"2019","title":"Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2019","url":"https://ijcai.org/proceedings/2019/","year":2019,"location":"Macao","dates":"10-16 August 2019","editors":"Sarit Kraus","isbn":"978-0-9992411-4-1","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2019"},"section":"Machine Learning A-L","track":"Main track","number":368,"pages":{"first":2649,"last":2655},"keywords":["Machine Learning: Learning Theory","Machine Learning: Unsupervised Learning","Computer Vision: Statistical Methods and Machine Learning","Machine Learning Applications: Applications of Unsupervised Learning"],"videos":[],"references":[{"position":1,"raw":"[Braman, 2010] Karen Braman. Third-order tensors as linear operators on a space of matrices. Linear Algebra & Its Applications, 433(7):1241-1253, 2010.","title":"Third-order tensors as linear operators on a space of matrices","authors":"Karen Braman","year":"2010","venue":"Linear Algebra & Its Applications","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Candés and Recht, 2009] Emmanuel J. Candés and Ben- jamin Recht. Exact matrix completion via convex op- timization. Foundations of Computational Mathematics, 9(6):717-772, 2009.","title":"Exact matrix completion via convex optimization","authors":"Emmanuel J Candés, Benjamin Recht","year":"2009","venue":"Foundations of Computational Mathematics","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Candés et al., 2011] Emmanuel J. Candés, Xiaodong Li, Yi Ma, and John Wright. Robust principal component analysis? Journal of the ACM, 58(3):1-73, May 2011.","title":"Robust principal component analysis","authors":"Emmanuel J Candés, Xiaodong Li, Yi Ma, John Wright","year":"2011","venue":"Journal of the ACM","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Chen, 2013] Yudong Chen. Incoherence-optimal matrix completion. IEEE TIT, 61(5):2909-2923, 2013.","title":"Incoherence-optimal matrix completion","authors":"Yudong Chen","year":"2013","venue":"IEEE TIT","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Dabov et al., 2007] Kostadin Dabov, Alessandro Foi, Vladimir Katkovnik, and Karen Egiazarian. Image restoration by sparse 3D transform-domain collaborative filtering. IEEE TIP, 16(8):2080-2095, August 2007.","title":"Image restoration by sparse 3D transform-domain collaborative filtering","authors":"Kostadin Dabov, Alessandro Foi, Vladimir Katkovnik, Karen Egiazarian","year":"2007","venue":"IEEE TIP","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Gandy et al., 2011] Silvia Gandy, Benjamin Recht, and Isao Yamada. Tensor completion and low-n-rank ten- sor recovery via convex optimization. Inverse Problems, 27(2):025010, 2011.","title":"Tensor completion and low-n-rank tensor recovery via convex optimization","authors":"Silvia Gandy, Benjamin Recht, Isao Yamada","year":"2011","venue":"Inverse Problems","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Gu et al., 2014] Quanquan Gu, Huan Gui, and Jiawei Han. Robust tensor decomposition with gross corruption. In Proceedings of the 27th Advances in Neural Information Processing Systems, pages 1422-1430, 2014.","title":"Robust tensor decomposition with gross corruption","authors":"Gu","year":"2014","venue":"Proceedings of the 27th Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Huang et al., 2014] Bo Huang, Cun Mu, Donald Goldfarb, and John Wright. Provable low-rank tensor recovery. Optimization-Online, page 4252, 2014.","title":"Provable low-rank tensor recovery","authors":"Bo Huang, Cun Mu, Donald Goldfarb, John Wright","year":"2014","venue":"Optimization-Online","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Jain and Oh, 2014] Prateek Jain and Sewoong Oh. Prov- able tensor factorization with missing data. In NIPS, pages 1431-1439, 2014.","title":"Provable tensor factorization with missing data","authors":"Prateek Jain, Sewoong Oh","year":"2014","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Karlsson et al., 2016] Lars Karlsson, Daniel Kressner, and André Uschmajew. Parallel algorithms for tensor com- pletion in the CP format. Parallel Comput., 57:222-234, September 2016.","title":"Parallel algorithms for tensor completion in the CP format","authors":"Karlsson","year":"2016","venue":"Parallel Comput","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Kilmer and Martin, 2011] Misha E. Kilmer and Carla D. Martin. Factorization strategies for third-order tensors. Linear Algebra and its Applications, 435(3):641-658, 2011.","title":"Factorization strategies for third-order tensors","authors":"Misha E Kilmer, Carla D Martin","year":"2011","venue":"Linear Algebra and its Applications","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Kilmer et al., 2013] Misha E. Kilmer, Karen Braman, Ning Hao, and Randy C. Hoover. Third-order tensors as oper- ators on matrices: a theoretical and computational frame- work with applications in imaging. SIAM Journal on Mat. Analy. & App., 34(1):148-172, 2013.","title":"Third-order tensors as operators on matrices: a theoretical and computational framework with applications in imaging","authors":"Misha E Kilmer, Karen Braman, Ning Hao, Randy C Hoover","year":"2013","venue":"SIAM Journal on Mat. Analy. & App","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Kokaram, 2004] Anil C. Kokaram. On missing data treat- ment for degraded video and film archives: A survey and a new bayesian approach. IEEE TIP, 13(3):397-415, March 2004.","title":"On missing data treatment for degraded video and film archives: A survey and a new bayesian approach","authors":"C Anil, Kokaram","year":"2004","venue":"IEEE TIP","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Kolda and Bader, 2009] Tamara G. Kolda and Brett W. Bader. Tensor decompositions and applications. SIAM Review, 51(3):455-500, 2009.","title":"Tensor decompositions and applications","authors":"Tamara G Kolda, Brett W Bader","year":"2009","venue":"SIAM Review","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Li, 2013] Xiaodong Li. Compressed sensing and matrix completion with constant proportion of corruptions. Con- structive Approximation, 37(1):73-99, 2013.","title":"Compressed sensing and matrix completion with constant proportion of corruptions","authors":"Xiaodong Li","year":"2013","venue":"Constructive Approximation","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Liu et al., 2013] Ji Liu, Przemyslaw Musialski, Peter Wonka, and Jieping Ye. Tensor completion for estimating missing values in visual data. IEEE T PAMI, 35(1):208- 220, 2013.","title":"Tensor completion for estimating missing values in visual data","authors":"Liu","year":"2013","venue":"IEEE T PAMI","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Lu et al., 2016] Canyi Lu, Jiashi Feng, Yudong Chen, Liu Wei, and Shuicheng Yan. Tensor robust principal compo- nent analysis: Exact recovery of corrupted low-rank ten- sors via convex optimization. In CVPR, pages 5249-5257, 2016.","title":"Tensor robust principal component analysis: Exact recovery of corrupted low-rank tensors via convex optimization","authors":"Lu","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Lu et al., 2018] Canyi Lu, Jiashi Feng, Yudong Chen, Liu Wei, Zhouchen Lin, and Shuicheng Yan. Tensor robust principal component analysis with a new tensor nuclear norm. arXiv:1804.03728, 2018.","title":"Tensor robust principal component analysis with a new tensor nuclear norm","authors":"Lu","year":"2018","venue":null,"doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Martin et al., 2001] David R. Martin, Charless Fowlkes, Doron Tal, and Jitendra Malik. A database of human seg- mented natural images and application to evaluating seg- mentation algorithms and measuring ecological statistics. ICCV, 2(11):416-423, 2001.","title":"A database of human segmented natural images and application to evaluating segmentation algorithms and measuring ecological statistics","authors":"David R Martin, Charless Fowlkes, Doron Tal, Jitendra Malik","year":"2001","venue":"ICCV","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Martin et al., 2013] Carla D. Martin, Richard Shafer, and Betsy Larue. An order-p tensor factorization with applica- tions in imaging. SIAM Journal on Scientific Computing, 35(1):A474-A490, 2013.","title":"An order-p tensor factorization with applications in imaging","authors":"Carla D Martin, Richard Shafer, Betsy Larue","year":"2013","venue":"SIAM Journal on Scientific Computing","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Mu et al., 2013] Cun Mu, Huang Bo, John Wright, and Donald Goldfarb. Square deal: Lower bounds and im- proved relaxations for tensor recovery. In ICML, pages 73-81, 2013.","title":"Square deal: Lower bounds and improved relaxations for tensor recovery","authors":"Cun Mu, Huang Bo, John Wright, Donald Goldfarb","year":"2013","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Shah et al., 2015] Parikshit Shah, Nikhil Rao, and Gongguo Tang. Sparse and low-rank tensor decomposition. In NIPS, pages 2548-2556, 2015.","title":"Sparse and low-rank tensor decomposition","authors":"Parikshit Shah, Nikhil Rao, Gongguo Tang","year":"2015","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Shang et al., 2014] Fanhua Shang, Yuanyuan Liu, James Cheng, and Hong Cheng. Robust principal component analysis with missing data. In CIKM, pages 1149-1158, November 2014.","title":"Robust principal component analysis with missing data","authors":"Shang","year":"2014","venue":"CIKM","doi":null,"url":null,"internal_url":null},{"position":24,"raw":"[Subrahmanyam et al., 2010] Gorthi R.K.S. Subrahmanyam, Ambasamudram N. Rajagopalan, and Rangarajan Ar- avind. Recursive framework for joint inpainting and de- noising of photographic films. Journal of the Optical So- ciety of America A, 27(5):1091-1099, 2010.","title":"Recursive framework for joint inpainting and denoising of photographic films","authors":"R K S Gorthi, Ambasamudram N Subrahmanyam, Rangarajan Rajagopalan, Aravind","year":"2010","venue":"Journal of the Optical Society of America A","doi":null,"url":null,"internal_url":null},{"position":25,"raw":"[Tomioka et al., 2010] Ryota Tomioka, Kohei Hayashi, and Hisashi Kashima. Estimation of low-rank tensors via con- vex optimization. arXiv:1010.0789, 2010.","title":"Estimation of low-rank tensors via convex optimization","authors":"Ryota Tomioka, Kohei Hayashi, Hisashi Kashima","year":"2010","venue":null,"doi":null,"url":null,"internal_url":null},{"position":26,"raw":"[Tomioka et al., 2011] Ryota Tomioka, Taiji Suzuki, Kohei Hayashi, and Hisashi Kashima. Statistical performance of convex tensor decomposition. In NIPS, pages 972-980, 2011.","title":"Statistical performance of convex tensor decomposition","authors":"Ryota Tomioka, Taiji Suzuki, Kohei Hayashi, Hisashi Kashima","year":"2011","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":27,"raw":"[Wright et al., 2013] John Wright, Arvind Ganesh, Kerui Min, and Yi Ma. Compressive principal component pur- suit. Information and Inference, 2(1):32-68, 2013.","title":"Compressive principal component pursuit","authors":"John Wright, Arvind Ganesh, Min Kerui, Yi Ma","year":"2013","venue":"Information and Inference","doi":null,"url":null,"internal_url":null},{"position":28,"raw":"[Zhang and Aeron, 2017] Zemin Zhang and Shuchin Aeron. Exact tensor completion using t-SVD. IEEE TSP, 65(6):1511-1526, 2017.","title":"Exact tensor completion using t-SVD","authors":"Aeron Zhang, Zemin Zhang, Shuchin Aeron","year":"2017","venue":"IEEE TSP","doi":null,"url":null,"internal_url":null},{"position":29,"raw":"[Zhang et al., 2014] Zemin Zhang, Gregory Ely, Shuchin Aeron, Ning Hao, and Misha Kilmer. Novel methods for multilinear data completion and de-noising based on tensor-SVD. In CVPR, pages 3842-3849, 2014.","title":"Novel methods for multilinear data completion and de-noising based on tensor-SVD","authors":"Zhang","year":"2014","venue":"CVPR","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}