Abstract
It is more and more common to encounter applications where the collected data is most naturally stored or represented in a multi-dimensional array, known as a tensor. The goal is often to approximate this tensor as a sum of some type of combination of basic elements, where the notation of what is a basic element is specific to the type of factorization employed. If the number of terms in the combination is few, the tensor factorization gives (implicitly) a sparse (approximate) representation of the data. The terms (e.g. vectors, matrices, tensors) in the combination themselves may also be sparse. This chapter highlights recent developments in the area of non-negative tensor factorization which admit such sparse representations. Specifically, we consider the approximate factorization of third and fourth order tensors into non-negative sums of types of outer-products of objects with one dimension less using the so-called t-product. A demonstration on an application in facial recognition shows the potential promise of the overall approach. We discuss a number of algorithmic options for solving the resulting optimization problems, and modification of such algorithms for increasing the sparsity.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
R. Harshman, ’70; J. Carroll and J. Chang, ’70.
- 2.
If \({{\fancyscript{A}}}\) is a fourth order tensor, the core tensor will be fourth order and there will be an additional summand and vector outer product.
References
Alex M, Alex OM (2002) Vasilescu, and Demetri Terzopoulos. Multilinear image analysis for facial recognition
Anderson DG (1965) Iterative procedures for nonlinear integral equations. J ACM 12(4):547–560
Andersson CA, Bro R (1998) Improving the speed of multi-way algorithms: part i. tucker3. Chemometr Intell Lab Syst 42:93–103
Arora S, Ge R, Kannan R, Moitra A (2012) Computing a nonnegative matrix factorization - provably. In: Proceedings of the 44th symposium on theory of computing, STOC ’12 ACM, New York, 145–162
Berry MW, Browne M,Langville AN, Paul Pauca V, Plemmons RJ (2006) Algorithms and applications for approximate nonnegative matrix factorization. In: Computational statistics and data analysis, pp 155–173
Boyd S. Sequential convex programming. Lecture slides
Braman K (2010) Third-order tensors as linear operators on a space of matrices. Linear Algebra Appl 433(7):1241–1253, Dec 2010
Bro R, De Jong S (1997) A fast non-negativity-constrained least squares algorithm. J Chemom 11:393–401
Chi EC, Kolda TG (2012) On tensors, sparsity, and nonnegative factorizations. http://arxiv.org/abs/1112.2414. August 2012 (AarXiv:1112.2414 [math.NA])
Cichocki A, Phan AH, Zdunek R, Amari S (2009) Nonnegative matrix and tensor factorizations: applications to exploratory multiway data analysis and blind source separation . Wiley, NJ (Preprint)
Cichocki A, Zdunek R, Choi S, Plemmons R, Amari S (2007) Nonnegative tensor factorization using alpha and beta divergencies. In: Proceedings of the 32nd international conference on acoustics, speech, and signal processing (ICASSP), Honolulu, April 2007
Cichocki A, Zdunek R, Choi S, Plemmons R, Amari S (2007) Novel multi-layer nonnegative tensor factorization with sparsity constraints. In: 8th international conference on adaptive and natural computing algorithms, Warsaw, MIT Press, Cambridge, April 2007
Donoho D, Stodden V (2003) When does non-negative matrix factorization give correct decomposition into parts?, p 2004. MIT Press, Cambridge
Elden L (2007) Matrix methods in data mining and pattern recognition. Publisher: society for industrial and applied mathematics (April 9 2007)
Friedlander MP, Hatz K (2006) Computing nonnegative tensor factorizations. Technical Report TR-2006-21, Computer Science Department, University of British Columbia
Gillis N, Plemmons RJ (2013) Sparse nonnegative matrix underapproximation and its application to hyperspectral image analysis. Linear Alg Appl 438(10):3991–4007
Gleich DF, Greif C, Varah JM (2012)The power and Arnoldi methods in an algebra of circulants. Numer. Linear Algebra Appl.2012. Published online in Wiley Online Library (www.wileyonlinelibrary.com). doi:10.1002/nla.1845
Hao N, Kilmer ME, Braman K, Hoover RC (2013) Facial recognition using tensor–tensor decompositions, SIAM J Imaging Sci 6(1):437–463
Homer Peng Ni, Walker F (July 2011) Anderson acceleration for fixed-point iterations. SIAM J Numer Anal 49(4):1715–1735
Kilmer ME, Braman K, Hao N, Hoover RC (2013) Third order tensors as operators on matrices: a theoretical and computational framework with applications in imaging. SIAM J Matrix Anal Appl 34(1):148–172
Kilmer ME, Kolda TG (2011) Approximations of third order tensors as sums of (non-negative) low-rank product cyclic tensors. In: Householder Symposium XVIII Plenary Talk. (there seems to be no live page with the book of abstracts), June 2011
Kilmer ME, MartinCarla D, Perrone L (2008) A third-order generalization of the matrix SVD as a product of third-order tensors. Technical report TR-2008-4, Department of Computer Science, Tufts University
Kilmer ME, Martin CD (2011) Factorization strategies for third-order tensors. Linear Algebra Appl 435(3):641–658. doi:10.1016/j.laa.2010.09.020 (Special Issue in Honor of Stewart’s GW 70th birthday)
Kilmer ME, Martin CD, Perrone L (2008) A third-order generalization of the matrix svd as a product of third-order tensors. In: Tufts computer science technical, report, 10 2008
Kolda TG, Bader BW (2009) Tensor decompositions and applications. SIAM Rev 51(3):455–500
Kruskal JB (1989) Rank, decomposition, and uniqueness for 3-way and \(n\)-way arrays. In: Coppi R, Bolasco S (eds) Multiway data analysis. Elsevier, Amsterdam, pp 7–18
Kumar A, Sindhwani V, Kambadur P (2012) Fast conical hull algorithms for near-separable non-negative matrix factorization. arXiv:1210.1190 [stat.ML], Oct 2012
Lee D, Seung H (1999) Learning the parts of objects by non-negative matrix factorization. Nature 401:788791
Lin Chih-Jen (2007) On the convergence of multiplicative update algorithms for nonnegative matrix factorization. IEEE Trans Neural Netw 18(6):1589–1596
Martin CD, Shafer R, LaRue B (2011) A recursive idea for multiplying order-p tensors. SIAM J Sci Comput (submitted July 2011)
Nocedal J, Wright SJ (2006) Numerical optimization, 2nd edn. Springer, New York
Vasilescu MAO, Terzopoulos D(2002) Multilinear analysis of image ensembles: Tensorfaces. In: Proceedings of the 7th European conference on computer vision ECCV 2002. Lecture notes in computer science, Vol 2350, 447–460
Vasilescu MAO, Terzopoulos D (2002) Multilinear image analysis for face recognition. In: Proceedings of the International conference on patter recognition ICPR 2002, vol 2. Quebec City, pp 511–514
Vasilescu MAO, Terzopoulos D (2003) Multilinear subspace analysis of image ensembles. In: Proceedings of the 2003 IEEE Computer society conference on computer vision and pattern recognition CVPR 2003, 93–99
Sylvestre EA, Lawton WH (1971) Self modeling curve resolution. Technometrics 13(3):617–633
Acknowledgments
We would like to thank Homer F. Walker from the Mathematical Science Department, Worcester Polytechnic Institute for his assistant with Anderson Acceleration. We would also like to acknowledge Hao and Kilmer’s support from NSF-DMS 0914957.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Hao, N., Horesh, L., Kilmer, M.E. (2014). Nonnegative Tensor Decomposition. In: Carmi, A., Mihaylova, L., Godsill, S. (eds) Compressed Sensing & Sparse Filtering. Signals and Communication Technology. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-38398-4_5
Download citation
DOI: https://doi.org/10.1007/978-3-642-38398-4_5
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-38397-7
Online ISBN: 978-3-642-38398-4
eBook Packages: EngineeringEngineering (R0)