English

Spectral norm of a symmetric tensor and its computation

Optimization and Control 2020-01-17 v3 Mathematical Physics math.MP

Abstract

We show that the spectral norm of a dd-mode real or complex symmetric tensor in nn variables can be computed by finding the fixed points of the corresponding polynomial map. For a generic complex symmetric tensor the number of fixed points is finite, and we give upper and lower bounds for the number of fixed points. For n=2n=2 we show that these fixed points are the roots of a corresponding univariate polynomial of degree at most (d1)2+1(d-1)^2+1, except certain cases, which are completely analyzed. In particular, for n=2n=2 the spectral norm of dd-symmetric tensor is polynomially computable in dd with a given relative precision. For a fixed n>2n>2 we show that the spectral norm of a dd-mode symmetric tensor is polynomially computable in dd with a given relative precision with respect to the Hilbert-Schmidt norm of the tensor. These results show that the geometric measure of entanglement of dd-mode symmetric qunits on Cn\mathbb{C}^n are polynomially computable for a fixed nn.

Keywords

Cite

@article{arxiv.1808.03864,
  title  = {Spectral norm of a symmetric tensor and its computation},
  author = {Shmuel Friedland and Li Wang},
  journal= {arXiv preprint arXiv:1808.03864},
  year   = {2020}
}

Comments

31 pages. arXiv admin note: substantial text overlap with arXiv:1608.01354, to appear in Mathematics of Computation, AMS