Robust matrix completion with complex noise

  • Li TangEmail author
  • Weili Guan


Matrix completion plays an important role in machine learning and data mining. Although a great number of algorithms have been developed for this issue, most of them can cope with only the Gaussian noise or sparse outliers. This paper focus on an intractable setting that the known entries are corrupted by Gaussian noise and sparse outliers simultaneously. Specifically, we construct a novel model with a loss function derived from the celebrated Huber function. Furthermore, an efficient optimization method is presented to solve the constructed model. The promising performance of our algorithm is demonstrated via numerous experiments on several benchmark datasets.


Matrix completion Subspace learning Sparse outliers 



  1. 1.
    Agudo A, Pijoan M, Moreno-Noguer F (2018) Image collection pop-up: 3d reconstruction and clustering of rigid and non-rigid categories. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 2607–2615Google Scholar
  2. 2.
    Basri R, Jacobs D, Kemelmacher I (2007) Photometric stereo with general, unknown lighting. Int J Comput Vis 72(3):239–257CrossRefGoogle Scholar
  3. 3.
    Cai J-F, Candès EJ, Shen Z (2010) A singular value thresholding algorithm for matrix completion. SIAM J Optim 20(4):1956–1982MathSciNetCrossRefGoogle Scholar
  4. 4.
    Cambier L, Absil P-A (2016) Robust low-rank matrix completion by riemannian optimization. SIAM J Sci Comput 38(5):S440–S460MathSciNetCrossRefGoogle Scholar
  5. 5.
    Candès EJ, Li X, Ma Y, Wright J (2011) Robust principal component analysis? J ACM (JACM) 58(3):11MathSciNetCrossRefGoogle Scholar
  6. 6.
    Candès EJ, Tao T (2010) The power of convex relaxation: near-optimal matrix completion. IEEE Trans Inf Theory 56(5):2053–2080MathSciNetCrossRefGoogle Scholar
  7. 7.
    Dai Y, Li H, He M (2014) A simple prior-free method for non-rigid structure-from-motion factorization. Int J Comput Vis 107(2):101–122MathSciNetCrossRefGoogle Scholar
  8. 8.
    Elsener A, Van De Geer S (2016) Robust low-rank matrix estimationGoogle Scholar
  9. 9.
    Eriksson A, Van Den Hengel A (2010) Efficient computation of robust low-rank matrix approximations in the presence of missing data using the l1 normGoogle Scholar
  10. 10.
    Goldberg K, Roeder T, Gupta D, Perkins C (2001) Eigentaste: a constant time collaborative filtering algorithm. Information Retrieval 4(2):133–151CrossRefGoogle Scholar
  11. 11.
    Gu S, et al. (2017) Weighted nuclear norm minimization and its applications to low level vision. Int J Comput Vis 121(2):183–208CrossRefGoogle Scholar
  12. 12.
    He J, Balzano L, Szlam A (2012) Incremental gradient on the grassmannian for online foreground and background separation in subsampled video. In: 2012 IEEE conference on computer vision and pattern recognition (CVPR). IEEE, pp 1568–1575Google Scholar
  13. 13.
    Herlocker JL, Konstan JA, Borchers A, Riedl J (2017) An algorithmic framework for performing collaborative filtering. In: ACM SIGIR forum, vol 51. ACM, pp 227–234Google Scholar
  14. 14.
    Hu Y, Zhang D, Ye J, Li X, He X (2013) Fast and accurate matrix completion via truncated nuclear norm regularization. IEEE Trans Pattern Anal Mach Intell 35(9):2117–2130CrossRefGoogle Scholar
  15. 15.
    Hu Z, Nie F, Wang R, Li X (2019) Multi-view spectral clustering via integrating nonnegative embedding and spectral embedding. Inf FusionGoogle Scholar
  16. 16.
    Huber PJ (2011) Robust statistics. In: International encyclopedia of statistical science. Springer, pp 1248–1251Google Scholar
  17. 17.
    Jegou H, Douze M, Schmid C (2008) Hamming embedding and weak geometric consistency for large scale image search. In: European conference on computer vision. Springer, pp 304–317Google Scholar
  18. 18.
    Ke Q, Kanade T (2005) Robust 1 norm factorization in the presence of outliers and missing data by alternative convex programming. In: IEEE computer society conference on computer vision and pattern recognition, 2005. CVPR 2005, vol 1. IEEE, pp 739–746Google Scholar
  19. 19.
    Lin Z, Xu C, Zha H (2018) Robust matrix factorization by majorization minimization. IEEE Trans Pattern Anal Mach Intell 40(1):208–220CrossRefGoogle Scholar
  20. 20.
    Liu G, et al. (2013) Robust recovery of subspace structures by low-rank representation. IEEE Trans Pattern Anal Mach Intell 35(1):171–184CrossRefGoogle Scholar
  21. 21.
    Liu Q, Davoine F, Yang J, Cui Y, Jin Z, Han F (2019) A fast and accurate matrix completion method based on QR decomposition and 2,1 -norm minimization. IEEE Trans Neural Netw Learning Syst 30(3):803–817MathSciNetCrossRefGoogle Scholar
  22. 22.
    Lu C, Tang J, Yan S, Lin Z (2014) Generalized nonconvex nonsmooth low-rank minimization. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 4130–4137Google Scholar
  23. 23.
    Lu C, Zhu C, et al. (2015) Generalized singular value thresholding. In: AAAI, pp 1805–1811Google Scholar
  24. 24.
    Meng D, Xu Z, Zhang L, Ji Z (2013) A cyclic weighted median method for l1 low-rank matrix factorization with missing entries. In: AAAI, vol 4, p 6Google Scholar
  25. 25.
    Ngo T, Saad Y (2012) Scaled gradients on grassmann manifolds for matrix completion. In: Advances in neural information processing systems, pp 1412–1420Google Scholar
  26. 26.
    Nie F, Hu Z, Li X (2018) Calibrated multi-task learning. In: Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining . ACM, pp 2012–2021Google Scholar
  27. 27.
    Nie F, Hu Z, Li X (2019) Matrix completion based on non-convex low-rank approximation. IEEE Trans Image Process 28(5):2378–2388MathSciNetCrossRefGoogle Scholar
  28. 28.
    Nie F, Huang H, Ding CHQ (2012) Low-rank matrix recovery via efficient schatten p-norm minimization. In: AAAIGoogle Scholar
  29. 29.
    Philbin J, Chum O, Isard M, Sivic J, Zisserman A (2007) Object retrieval with large vocabularies and fast spatial matching. In: IEEE conference on computer vision and pattern recognition, 2007. CVPR’07. IEEE, p 2007Google Scholar
  30. 30.
    Pierre A, Vincent C, Guillaume L (2019) Estimation bounds and sharp oracle inequalities of regularized procedures with lipschitz loss functions. The Annals of Statistics 47(4):2117–2144MathSciNetCrossRefGoogle Scholar
  31. 31.
    Kwok J, Yao QM (2018) Scalable robust matrix factorization with nonconvex loss. In: Proceedings of the advances in neural information processing systems, pp 369–404Google Scholar
  32. 32.
    Shang F, et al. (2017) Bilinear factor matrix norm minimization for robust pca: algorithms and applications. IEEE Trans Pattern Anal Mach Intell (1):1–1Google Scholar
  33. 33.
    Shang F, Liu Y, Cheng J, Yan D (2018) Fuzzy double trace norm minimization for recommendation systems. IEEE Trans Fuzzy Systems 26(4):2039–2049CrossRefGoogle Scholar
  34. 34.
    Shang F, Liu Y, Tong H, Cheng J, Cheng H (2015) Robust bilinear factorization with missing and grossly corrupted observations. Inf Sci 307:53–72MathSciNetCrossRefGoogle Scholar
  35. 35.
    Sun Q, Xiang S, Ye J (2013) Robust principal component analysis via capped norms. In: Proceedings of the 19th ACM SIGKDD international conference on knowledge discovery and data mining. ACM, pp 311–319Google Scholar
  36. 36.
    Tanner J, Ke W (2016) Low rank matrix completion by alternating steepest descent methods. Appl Comput Harmon Anal 40(2):417–429MathSciNetCrossRefGoogle Scholar
  37. 37.
    Tomasi C, Kanade T (1992) Shape and motion from image streams under orthography: a factorization method. Int J Comput Vis 9(2):137–154CrossRefGoogle Scholar
  38. 38.
    Vandereycken B (2013) Low-rank matrix completion by riemannian optimization. SIAM J Optim 23(2):1214–1236MathSciNetCrossRefGoogle Scholar
  39. 39.
    Wang Y, Wu L, Lin X, Gao J (2018) Multiview spectral clustering via structured low-rank matrix factorization. IEEE Trans Neural Netw Learning Syst 29 (10):4833–4843CrossRefGoogle Scholar
  40. 40.
    Wen Z, Yin W, Zhang Y (2012) Solving a low-rank factorization model for matrix completion by a nonlinear successive over-relaxation algorithm. Math Program Comput 4(4):333–361MathSciNetCrossRefGoogle Scholar
  41. 41.
    Yao Q (2018) Scalable tensor completion with nonconvex regularization. arXiv:1807.08725
  42. 42.
    Yao Q, Kwok JT, Wang T, Liu T-Y (2017) Large-scale low-rank matrix learning with nonconvex regularizers. arXiv:1708.00146
  43. 43.
    Zhao L, Babu P, Palomar DP (2016) Efficient algorithms on robust low-rank matrix completion against outliers. IEEE Trans Signal Processing 64(18):4767–4780MathSciNetCrossRefGoogle Scholar
  44. 44.
    Zheng Y, Liu G, Sugimoto S, Yan S, Okutomi M (2012) Practical low-rank matrix approximation under robust l 1-norm. In: 2012 IEEE conference on computer vision and pattern recognition (CVPR). IEEE, p 2012Google Scholar
  45. 45.
    Zhu L, Zi H, Li Z, Xie L, Shen HT (2018) Exploring auxiliary context: discrete semantic transfer hashing for scalable image retrieval. IEEE Trans Neural Netw Learning Syst 29(11):5264–5276MathSciNetCrossRefGoogle Scholar
  46. 46.
    Zhu L, Zi H, Liu X, He X, Sun J, Zhou X (2017) Discrete multimodal hashing with canonical views for robust mobile landmark search. IEEE Trans Multimedia 19(9):2066–2079CrossRefGoogle Scholar
  47. 47.
    Zhu L, Shen J, Xie L, Cheng Z (2017) Unsupervised visual hashing with semantic assistant for content-based image retrieval. IEEE Trans Knowl Data Eng 29 (2):472–486CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Teacher Education InstituteDaqing Normal UniversityDaqingChina
  2. 2.Hewlett Packard Enterprise SingaporeSingaporeSingapore

Personalised recommendations