Tensor convolutions and Hankel tensors

Abstract

Let A be an mth order n-dimensional tensor, where m, nare some positive integers and N:= m(n−1).Then A is called a Hankel tensor associated with a vector v\in {ℝ}^{N+\mathrm{1}} if A_{\sigma }=v_k for each k= 0, 1, …,Nwhenever σ= (i₁, …,i_m) satisfies i₁ +…+i_m = m+k.We introduce the elementary Hankel tensors which are some special Hankel tensors, and present all the eigenvalues of the elementary Hankel tensors for k= 0, 1, 2. We also show that a convolution can be expressed as the product of some third-order elementary Hankel tensors, and a Hankel tensor can be decomposed as a convolution of two Vandermonde matrices following the definition of the convolution of tensors. Finally, we use the properties of the convolution to characterize Hankel tensors and (0,1) Hankel tensors.

Keywords

Tensor / convolution / Hankel tensor / elementary Hankel tensor / symmetric tensor

References

[1] ChenY, QiL, WangQ. Computing extreme eigenvalues of large scale Hankel tensors. J Sci Comput, 2016, 68: 716–738

[2] ChenY, QiL,WangQ. Positive semi-definiteness and sum-of-squares property of fourth order four dimensional Hankel tensors.J Comput Appl Math, 2016, 302: 356–368

[3] ComonP, GolubG, LimL H, MourrainB. Symmetric tensors and symmetric tensor rank.SIAM J Matrix Anal Appl, 2008, 30: 1254–1279

[4] DingW, QiL, WeiY. Fast Hankel tensor-vector product and its application to exponential data fitting.Numer Linear Algebra Appl, 2015, 22: 814–832

[5] DingW, QiL, WeiY. Inheritance properties and sum-of-squares decomposition of Hankel tensors: theory and algorithms.BIT, 2017, 57: 169–190

[6] FazelM, PongT K, SunD, TsengP. Hankel matrix rank minimization with applications in system identification and realization.SIAM J Matrix Anal Appl, 2013, 34: 946–977

[7] HillarC J, LimL-H. Most tensor problems are NP-hard.J ACM, 2013, 60(6): 1–45

[8] LiG, QiL, WangQ. Positive semi-definiteness of generalized anti-circular tensors.Commun Math Sci, 2016, 14: 941–952

[9] LimL H. Singular values and eigenvalues of tensors: a variational approach. In: IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing. IEEE, 2006, 129–132

[10] OppenheimA V. Linear Time-Invariant Systems in Signals and Systems.2nd ed. Englewood: Prentice Hall, 1996

[11] PapyJ M, De LauauwerL, Van HuffelS. Exponential data fitting using multilinear algebra: The single-channel and multi-channel case.Numer Linear Algebra Appl, 2005, 12: 809–826

[12] QiL. Eigenvalues of a supersymmetric tensor and positive definiteness of an even degree multivariate form.Research Report, Department of Applied Mathematics, The Hong Kong Polytechnic University, 2004

[13] QiL. Eigenvalues of a real supersymmetric tensor.J Symbolic Comput, 2005, 40: 1302–1324

[14] QiL. Symmetric nonnegative tensors and copositive tensors.Linear Algebra Appl, 2013, 439: 228–238

[15] QiL. Hankel tensors: Associated Hankel matrices and Vandermonde decomposition.Commun Math Sci, 2015, 13: 113–125

[16] VarahJ M. Positive definite Hankel matrices of minimal condition.Linear Algebra Appl, 2003, 368: 303–314

[17] WangQ, LiG, QiL, XuY. New classes of positive semi-definite Hankel tensor. arXiv: 1411.2365v5

[18] XuC. Hankel tensors, Vandermonde tensors and their positivities.Linear Algebra Appl, 2016, 491: 56–72

RIGHTS & PERMISSIONS

2016 Higher Education Press and Springer-Verlag GmbH Germany