Abstract
We study symmetric tensor decompositions, i.e., decompositions of the form \(T = \sum _{i=1}^r u_i^{\otimes 3}\) where T is a symmetric tensor of order 3 and \(u_i \in \mathbb {C}^n\). In order to obtain efficient decomposition algorithms, it is necessary to require additional properties from the \(u_i\). In this paper we assume that the \(u_i\) are linearly independent. This implies \(r \le n\), i.e., the decomposition of T is undercomplete. We will moreover assume that \(r=n\) (we plan to extend this work to the case \(r<n\) in a forthcoming paper.) We give a randomized algorithm for the following problem: given T, an accuracy parameter \(\epsilon \), and an upper bound B on the condition number of the tensor, output vectors \(u'_i\) such that \(||u_i - u'_i|| \le \epsilon \) (up to permutation and multiplication by phases) with high probability. The main novel features of our algorithm are:
-
We provide the first algorithm for this problem that works in the computation model of finite arithmetic and requires only poly-logarithmic (in n, B and \(\frac{1}{\epsilon }\)) many bits of precision.
-
Moreover, this is also the first algorithm that runs in linear time in the size of the input tensor. It requires \(O(n^3)\) arithmetic operations for all accuracy parameters \(\epsilon = \frac{1}{\text {poly}(n)}\).
In order to obtain these results, we rely on a mix of techniques from algorithm design and algorithm analysis. The algorithm is a modified version of Jennrich’s algorithm for symmetric tensors. In terms of algorithm design, our main contribution lies in replacing the usual appeal to resolution of a linear system of equations [5, 12] by a matrix trace-based method. The analysis of the algorithm depends on the following components:
-
1.
We use the fast and numerically stable diagonalisation algorithm from [1]. We provide better guarantees for the approximate solution returned by the diagonalisation algorithm when the input matrix is diagonalisable.
-
2.
We show strong anti-concentration bounds for certain families of polynomials when the randomness is sampled uniformly from a discrete grid.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
\(\kappa (T)\) also appears in the sublinear term for the arithmetic complexity of the algorithm.
- 2.
CPD stands for Canonical Polyadic Decomposition, i.e., decomposition as a sum of rank-1 tensors.
References
Banks, J., Garza-Vargas, J., Kulkarni, A., Srivastava, N.: Pseudospectral shattering, the sign function, and diagonalization in nearly matrix multiplication time. In: IEEE 61st Annual Symposium on Foundations of Computer Science (FOCS) (2020). https://doi.org/10.1109/FOCS46700.2020.00056
Beltrán, C., Breiding, P., Vannieuwenhoven, N.: Pencil-based algorithms for tensor rank decomposition are not stable. SIAM J. Matrix Anal. Appl. 40(2), 739–773 (2019). https://doi.org/10.1137/18M1200531
Bernardi, A., Gimigliano, A., Idà, M.: Computing symmetric rank for symmetric tensors. J. Symb. Comput. 46(1), 34–53 (2011). https://doi.org/10.1016/j.jsc.2010.08.001
Bhargava, V., Saraf, S., Volkovich, I.: Reconstruction algorithms for low-rank tensors and depth-3 multilinear circuits. In: Proceedings of the 53rd Annual ACM SIGACT Symposium on Theory of Computing, pp. 809–822 (2021). https://doi.org/10.1145/3406325.3451096
Bhaskara, A., Charikar, M., Moitra, A., Vijayaraghavan, A.: Smoothed analysis of tensor decompositions. In: Proceedings of the Forty-Sixth Annual ACM Symposium on Theory of Computing (STOC), pp. 594–603 (2014). https://doi.org/10.1145/2591796.2591881
Brachat, J., Comon, P., Mourrain, B., Tsigaridas, E.P.: Symmetric tensor decomposition. In: 17th European Signal Processing Conference (EUSIPCO). IEEE (2009). https://ieeexplore.ieee.org/document/7077748/
Forbes, M.A., Shpilka, A.: A PSPACE construction of a hitting set for the closure of small algebraic circuits. In: Proceedings of the 50th Annual ACM SIGACT Symposium on Theory of Computing (STOC), pp. 1180–1192 (2018). https://doi.org/10.1145/3188745.3188792
Goyal, N., Vempala, S., Xiao, Y.: Fourier PCA and robust tensor decomposition. In: Proceedings of the Forty-Sixth Annual ACM Symposium on Theory of Computing (STOC), pp. 584–593 (2014). https://doi.org/10.1145/2591796.2591875
Harshman, R.: Foundations of the PARAFAC procedure: models and conditions for an "explanatory" multi-mode factor analysis. UCLA Working Papers in Phonetics (1970). https://www.psychology.uwo.ca/faculty/harshman/wpppfac0.pdf
Håstad, J.: Tensor rank is NP-complete. In: Ausiello, G., Dezani-Ciancaglini, M., Della Rocca, S.R. (eds.) ICALP 1989. LNCS, vol. 372, pp. 451–460. Springer, Heidelberg (1989). https://doi.org/10.1007/BFb0035776
Higham, N.J.: Accuracy and Stability of Numerical Algorithms. Society for Industrial and Applied Mathematics, 2nd edn. (2002). https://doi.org/10.1137/1.9780898718027
Kayal, N.: Efficient algorithms for some special cases of the polynomial equivalence problem. In: Proceedings of the Twenty-Second Annual ACM-SIAM Symposium on Discrete Algorithms (SODA), pp. 1409–1421 (2011). https://doi.org/10.1137/1.9781611973082.108
Koiran, P., Saha, S.: Complete decomposition of symmetric tensors in linear time and polylogarithmic precision (2022). https://doi.org/10.48550/ARXIV.2211.07407
Kolda, T.G., Bader, B.W.: Tensor decompositions and applications. SIAM Rev. 51(3), 455–500 (2009). https://doi.org/10.1137/07070111X
Kruskal, J.B.: Three-way arrays: rank and uniqueness of trilinear decompositions, with application to arithmetic complexity and statistics. Linear Algebra Appl. 18(2), 95–138 (1977). https://doi.org/10.1016/0024-3795(77)90069-6
Leurgans, S.E., Ross, R.T., Abel, R.B.: A decomposition for three-way arrays. SIAM J. Matrix Anal. Appl. 14(4), 1064–1083 (1993). https://doi.org/10.1137/0614071
Moitra, A.: Algorithmic Aspects of Machine Learning. Cambridge University Press, Cambridge (2018) https://books.google.fr/books?id=ruVqDwAAQBAJ
Peleg, S., Shpilka, A., Volk, B.L.: Tensor reconstruction beyond constant rank (2022). https://doi.org/10.48550/ARXIV.2209.04177
Shitov, Y.: How hard is the tensor rank? (2016). https://doi.org/10.48550/ARXIV.1611.01559
Smale, S.: Complexity theory and numerical analysis. Acta Numer 6, 523–551 (1997). https://doi.org/10.1017/S0962492900002774
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Koiran, P., Saha, S. (2023). Complete Decomposition of Symmetric Tensors in Linear Time and Polylogarithmic Precision. In: Mavronicolas, M. (eds) Algorithms and Complexity. CIAC 2023. Lecture Notes in Computer Science, vol 13898. Springer, Cham. https://doi.org/10.1007/978-3-031-30448-4_22
Download citation
DOI: https://doi.org/10.1007/978-3-031-30448-4_22
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-30447-7
Online ISBN: 978-3-031-30448-4
eBook Packages: Computer ScienceComputer Science (R0)