Multilinear subspace learning

Multilinear subspace learning is an approach for

disentangling the causal factor of data formation and
performing dimensionality reduction.[1][2][3][4][5] The
Dimensionality reduction can be performed on a data
tensor that contains a collection of observations have
been vectorized,[1] or observations that are treated as
matrices and concatenated into a data tensor.[6][7] Here
are some examples of data tensors whose observations
are vectorized or whose observations are matrices
concatenated into data tensor images (2D/3D), video
sequences (3D/4D), and hyperspectral cubes (3D/4D).

The mapping from a high-dimensional vector space to

a set of lower dimensional vector spaces is a
multilinear projection.[4] When observations are A video or an image sequence represented as a
retained in the same organizational structure as third-order tensor of column x row x time for
matrices or higher order tensors, their representations multilinear subspace learning.
are computed by performing linear projections into the
column space, row space and fiber space.[6]

Multilinear subspace learning algorithms are higher-order generalizations of linear subspace learning
methods such as principal component analysis (PCA), independent component analysis (ICA), linear
discriminant analysis (LDA) and canonical correlation analysis (CCA).

Background
Multilinear methods may be causal in nature and perform causal inference, or they may be simple
regression methods from which no causal conclusion are drawn.

Linear subspace learning algorithms are traditional dimensionality reduction techniques that are well suited
for datasets that are the result of varying a single causal factor. Unfortunately, they often become inadequate
when dealing with datasets that are the result of multiple causal factors. .

Multilinear subspace learning can be applied to observations whose measurements were vectorized and
organized into a data tensor for causally aware dimensionality reduction.[1] These methods may also be
employed in reducing horizontal and vertical redundancies irrespective of the causal factors when the
observations are treated as a "matrix" (ie. a collection of independent column/row observations) and
concatenated into a tensor.[8][9]

Algorithms

Multilinear principal component analysis

Historically, multilinear principal component analysis has been referred to as "M-mode PCA", a
terminology which was coined by Peter Kroonenberg.[10] In 2005, Vasilescu and Terzopoulos introduced
the Multilinear PCA[11] terminology as a way to better differentiate between multilinear tensor
decompositions that computed 2nd order statistics associated with each data tensor mode,[1][2][3][12][13] and
subsequent work on Multilinear Independent Component Analysis[11] that computed higher order statistics
for each tensor mode. MPCA is an extension of PCA.

Multilinear independent component analysis

Multilinear independent component analysis[11] is an extension of ICA.

Multilinear linear discriminant analysis

Multilinear extension of LDA

TTP-based: Discriminant Analysis with Tensor Representation (DATER)[9]

TTP-based: General tensor discriminant analysis (GTDA)[14]
TVP-based: Uncorrelated Multilinear Discriminant Analysis (UMLDA)[15]

Multilinear canonical correlation analysis

Multilinear extension of CCA

TTP-based: Tensor Canonical Correlation Analysis (TCCA)[16]

TVP-based: Multilinear Canonical Correlation Analysis (MCCA)[17]
TVP-based: Bayesian Multilinear Canonical Correlation Analysis (BMTF)[18]
A TTP is a direct projection of a high-dimensional tensor to a low-dimensional tensor of the
same order, using N projection matrices for an Nth-order tensor. It can be performed in N
steps with each step performing a tensor-matrix multiplication (product). The N steps are
exchangeable.[19] This projection is an extension of the higher-order singular value
decomposition[19] (HOSVD) to subspace learning.[13] Hence, its origin is traced back to the
Tucker decomposition[20] in 1960s.

A TVP is a direct projection of a high-dimensional tensor to a low-dimensional vector, which

is also referred to as the rank-one projections. As TVP projects a tensor to a vector, it can be
viewed as multiple projections from a tensor to a scalar. Thus, the TVP of a tensor to a P-
dimensional vector consists of P projections from the tensor to a scalar. The projection from
a tensor to a scalar is an elementary multilinear projection (EMP). In EMP, a tensor is
projected to a point through N unit projection vectors. It is the projection of a tensor on a
single line (resulting a scalar), with one projection vector in each mode. Thus, the TVP of a
tensor object to a vector in a P-dimensional vector space consists of P EMPs. This
projection is an extension of the canonical decomposition,[21] also known as the parallel
factors (PARAFAC) decomposition.[22]

Typical approach in MSL

There are N sets of parameters to be solved, one in each mode. The solution to one set often depends on the
other sets (except when N=1, the linear case). Therefore, the suboptimal iterative procedure in[23] is
followed.

1. Initialization of the projections in each mode

2. For each mode, fixing the projection in all the other mode, and solve for the projection in the
current mode.
3. Do the mode-wise optimization for a few iterations or until convergence.

This is originated from the alternating least square method for multi-way data analysis.[10]

Code
MATLAB Tensor Toolbox (https://web.archive.org/web/20110717172720/http://csmr.ca.sandi
a.gov/~tgkolda/TensorToolbox/) by Sandia National Laboratories.
The MPCA algorithm written in Matlab (MPCA+LDA included) (http://www.mathworks.com/m
atlabcentral/fileexchange/26168).
The UMPCA algorithm written in Matlab (data included) (http://www.mathworks.com/matlabc
entral/fileexchange/35432).
The UMLDA algorithm written in Matlab (data included) (http://www.mathworks.fr/matlabcentr
al/fileexchange/35782).

Tensor data sets

3D gait data (third-order tensors): 128x88x20(21.2M) (http://www.dsp.utoronto.ca/~haiping/C
odeData/USFGait17_128x88x20.zip); 64x44x20(9.9M) (http://www.dsp.utoronto.ca/~haiping/
CodeData/USFGait17_64x44x20.zip); 32x22x10(3.2M) (http://www.dsp.utoronto.ca/~haipin
g/CodeData/USFGait17_32x22x10.zip);

See also
CP decomposition
Dimension reduction
Multilinear algebra
Multilinear Principal Component Analysis
Tensor
Tensor decomposition
Tensor software
Tucker decomposition

References
1. M. A. O. Vasilescu, D. Terzopoulos (2003) "Multilinear Subspace Analysis of Image
Ensembles" (http://www.cs.toronto.edu/~maov/tensorfaces/cvpr03.pdf), "Proceedings of the
IEEE Conference on Computer Vision and Pattern Recognition (CVPR’03), Madison, WI,
June, 2003"
 2. M. A. O. Vasilescu, D. Terzopoulos (2002) "Multilinear Analysis of Image Ensembles:
TensorFaces" (http://www.cs.toronto.edu/~maov/tensorfaces/Springer%20ECCV%202002_fi
les/eccv02proceeding_23500447.pdf), Proc. 7th European Conference on Computer Vision
(ECCV'02), Copenhagen, Denmark, May, 2002
3. M. A. O. Vasilescu,(2002) "Human Motion Signatures: Analysis, Synthesis, Recognition" (htt
p://www.media.mit.edu/~maov/motionsignatures/hms_icpr02_corrected.pdf), "Proceedings
of International Conference on Pattern Recognition (ICPR 2002), Vol. 3, Quebec City,
Canada, Aug, 2002, 456–460."
4. Vasilescu, M.A.O.; Terzopoulos, D. (2007). Multilinear Projection for Appearance-Based
Recognition in the Tensor Framework. IEEE 11th International Conference on Computer
Vision. pp. 1–8. doi:10.1109/ICCV.2007.4409067 (https://doi.org/10.1109%2FICCV.2007.44
09067)..
5. Lu, Haiping; Plataniotis, K.N.; Venetsanopoulos, A.N. (2013). Multilinear Subspace
Learning: Dimensionality Reduction of Multidimensional Data (http://www.crcpress.com/prod
uct/isbn/9781439857243). Chapman & Hall/CRC Press Machine Learning and Pattern
Recognition Series. Taylor and Francis. ISBN 978-1-4398572-4-3.
6. Lu, Haiping; Plataniotis, K.N.; Venetsanopoulos, A.N. (2011). "A Survey of Multilinear
Subspace Learning for Tensor Data" (http://www.dsp.utoronto.ca/~haiping/Publication/Surve
yMSL_PR2011.pdf) (PDF). Pattern Recognition. 44 (7): 1540–1551.
Bibcode:2011PatRe..44.1540L (https://ui.adsabs.harvard.edu/abs/2011PatRe..44.1540L).
doi:10.1016/j.patcog.2011.01.004 (https://doi.org/10.1016%2Fj.patcog.2011.01.004).
7. X. He, D. Cai, P. Niyogi, Tensor subspace analysis (http://books.nips.cc/papers/files/nips18/
NIPS2005_0249.pdf), in: Advances in Neural Information Processing Systemsc 18 (NIPS),
2005.
8. "Future Directions in Tensor-Based Computation and Modeling" (http://www.cs.cornell.edu/c
v/tenwork/finalreport.pdf) (PDF). May 2009.
9. S. Yan, D. Xu, Q. Yang, L. Zhang, X. Tang, and H.-J. Zhang, "Discriminant analysis with
tensor representation (http://portal.acm.org/citation.cfm?id=1068959)," in Proc. IEEE
Conference on Computer Vision and Pattern Recognition, vol. I, June 2005, pp. 526–532.
10. P. M. Kroonenberg and J. de Leeuw, Principal component analysis of three-mode data by
means of alternating least squares algorithms (https://doi.org/10.1007%2FBF02293599),
Psychometrika, 45 (1980), pp. 69–97.
11. M. A. O. Vasilescu, D. Terzopoulos (2005) "Multilinear Independent Component Analysis" (ht
tp://www.media.mit.edu/~maov/mica/mica05.pdf), "Proceedings of the IEEE Conference on
Computer Vision and Pattern Recognition (CVPR’05), San Diego, CA, June 2005, vol.1,
547–553."
12. M.A.O. Vasilescu, D. Terzopoulos (2004) "TensorTextures: Multilinear Image-Based
Rendering", M. A. O. Vasilescu and D. Terzopoulos, Proc. ACM SIGGRAPH 2004
Conference Los Angeles, CA, August, 2004, in Computer Graphics Proceedings, Annual
Conference Series, 2004, 336–342. (http://www.media.mit.edu/~maov/tensortextures/Vasiles
cu_siggraph04.pdf)
13. H. Lu, K. N. Plataniotis, and A. N. Venetsanopoulos, "MPCA: Multilinear principal
component analysis of tensor objects (https://dx.doi.org/10.1109/TNN.2007.901277)," IEEE
Trans. Neural Netw., vol. 19, no. 1, pp. 18–39, January 2008.
14. D. Tao, X. Li, X. Wu, and S. J. Maybank, "General tensor discriminant analysis and gabor
features for gait recognition (https://dx.doi.org/10.1109/TPAMI.2007.1096)," IEEE Trans.
Pattern Anal. Mach. Intell., vol. 29, no. 10, pp. 1700–1715, October 2007.
15. H. Lu, K. N. Plataniotis, and A. N. Venetsanopoulos, "Uncorrelated multilinear discriminant
analysis with regularization and aggregation for tensor object recognition (https://dx.doi.org/1
0.1109/TNN.2008.2004625)," IEEE Trans. Neural Netw., vol. 20, no. 1, pp. 103–123,
January 2009.
16. T.-K. Kim and R. Cipolla. "Canonical correlation analysis of video volume tensors for action
categorization and detection (http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=454
7427)," IEEE Trans. Pattern Anal. Mach. Intell., vol. 31, no. 8, pp. 1415–1428, 2009.
17. H. Lu, "Learning Canonical Correlations of Paired Tensor Sets via Tensor-to-Vector
Projection (http://www.dsp.utoronto.ca/~haiping/Publication/MCCA_IJCAI2013.pdf),"
Proceedings of the 23rd International Joint Conference on Artificial Intelligence (IJCAI 2013),
Beijing, China, August 3–9, 2013.
18. Khan, Suleiman A.; Kaski, Samuel (2014-09-15). Calders, Toon; Esposito, Floriana;
Hüllermeier, Eyke; Meo, Rosa (eds.). Machine Learning and Knowledge Discovery in
Databases. Lecture Notes in Computer Science. Springer Berlin Heidelberg. pp. 656–671.
doi:10.1007/978-3-662-44848-9_42 (https://doi.org/10.1007%2F978-3-662-44848-9_42).
ISBN 9783662448472.
19. L.D. Lathauwer, B.D. Moor, J. Vandewalle, A multilinear singular value decomposition (http://
portal.acm.org/citation.cfm?id=354398), SIAM Journal of Matrix Analysis and Applications
vol. 21, no. 4, pp. 1253–1278, 2000
20. Ledyard R Tucker (September 1966). "Some mathematical notes on three-mode factor
analysis". Psychometrika. 31 (3): 279–311. doi:10.1007/BF02289464 (https://doi.org/10.100
7%2FBF02289464). PMID 5221127 (https://pubmed.ncbi.nlm.nih.gov/5221127).
S2CID 44301099 (https://api.semanticscholar.org/CorpusID:44301099).
21. J. D. Carroll & J. Chang (1970). "Analysis of individual differences in multidimensional
scaling via an n-way generalization of 'Eckart–Young' decomposition". Psychometrika. 35
(3): 283–319. doi:10.1007/BF02310791 (https://doi.org/10.1007%2FBF02310791).
S2CID 50364581 (https://api.semanticscholar.org/CorpusID:50364581).
22. R. A. Harshman, Foundations of the PARAFAC procedure: Models and conditions for an
"explanatory" multi-modal factor analysis (http://publish.uwo.ca/~harshman/wpppfac0.pdf)
Archived (https://web.archive.org/web/20041010092429/http://publish.uwo.ca/~harshman/w
pppfac0.pdf) 2004-10-10 at the Wayback Machine. UCLA Working Papers in Phonetics, 16,
pp. 1–84, 1970.
23. L. D. Lathauwer, B. D. Moor, J. Vandewalle, On the best rank-1 and rank-(R1, R2, ..., RN )
approximation of higher-order tensors (http://portal.acm.org/citation.cfm?id=354405), SIAM
Journal of Matrix Analysis and Applications 21 (4) (2000) 1324–1342.

Retrieved from "https://en.wikipedia.org/w/index.php?title=Multilinear_subspace_learning&oldid=1164373209"