CANDECOMP/PARAFAC Decomposition of High-Order Tensors Through Tensor Reshaping
作者： Anh-Huy Phan , Petr Tichavsky , Andrzej Cichocki
关键词: Mathematics 、 Algebra 、 Matrix decomposition 、 Symmetric tensor 、 Kruskal's algorithm 、 Cartesian tensor 、 Tensor contraction 、 Basis (linear algebra) 、 Tensor 、 Multilinear subspace learning
摘要: In general, algorithms for order-3 CANDECOMP/ PARAFAC (CP), also coined canonical polyadic decomposition (CPD), are easy to implement and can be extended higher order CPD. Unfortunately, the become computationally demanding, they often not applicable relatively large scale tensors. this paper, by exploiting uniqueness of CPD relation a tensor in Kruskal form its unfolded tensor, we propose fast approach deal with problem. Instead directly factorizing high data method decomposes an lower order, e.g., tensor. On basis estimated structured same dimension as is then generated, decomposed find final solution using addition, strategies unfold tensors suggested practically verified paper.
harvard.edu
arxiv.org
sci-hub.se 参考文章(80)
Morten Mørup, Lars Kai Hansen, Christoph S. Herrmann, Josef Parnas, Sidse M. Arnfred, Parallel Factor Analysis as an exploratory tool for wavelet transformed event-related EEG. NeuroImage. ,vol. 29, pp. 938- 947 ,(2006) , 10.1016/J.NEUROIMAGE.2005.08.005
Henk A.L. Kiers, Richard A. Harshman, Relating two proposed methods for speedup of algorithms for fitting two- and three-way principal component and related multilinear models Chemometrics and Intelligent Laboratory Systems. ,vol. 36, pp. 31- 40 ,(1997) , 10.1016/S0169-7439(96)00074-3
Anh Huy Phan, Andrzej Cichocki, Tensor decompositions for feature extraction and classification of high dimensional datasets Nonlinear Theory and Its Applications, IEICE. ,vol. 1, pp. 37- 68 ,(2010) , 10.1587/NOLTA.1.37
Kawee Numpacharoen, Amporn Atsawarungruangkit, Generating Correlation Matrices Based on the Boundaries of Their Coefficients PLoS ONE. ,vol. 7, pp. e48902- ,(2012) , 10.1371/JOURNAL.PONE.0048902
Karl Booksh, Bruce R. Kowalski, Error analysis of the generalized rank annihilation method Journal of Chemometrics. ,vol. 8, pp. 45- 63 ,(1994) , 10.1002/CEM.1180080106
I. V. Oseledets, E. E. Tyrtyshnikov, Breaking the Curse of Dimensionality, Or How to Use SVD in Many Dimensions SIAM Journal on Scientific Computing. ,vol. 31, pp. 3744- 3759 ,(2009) , 10.1137/090748330
Anders H Andersen, William S Rayens, Structure-seeking multilinear methods for the analysis of fMRI data NeuroImage. ,vol. 22, pp. 728- 739 ,(2004) , 10.1016/J.NEUROIMAGE.2004.02.026
J. Douglas Carroll, Jih-Jie Chang, Analysis of individual differences in multidimensional scaling via an n-way generalization of 'eckart-young' decomposition Psychometrika. ,vol. 35, pp. 283- 319 ,(1970) , 10.1007/BF02310791
Lek-Heng Lim, Pierre Comon, Blind Multilinear Identification IEEE Transactions on Information Theory. ,vol. 60, pp. 1260- 1280 ,(2014) , 10.1109/TIT.2013.2291876
Alwin Stegeman, Nicholas D. Sidiropoulos, On Kruskal's uniqueness condition for the Candecomp/Parafac decomposition Linear Algebra and its Applications. ,vol. 420, pp. 540- 552 ,(2007) , 10.1016/J.LAA.2006.08.010