English

On the Largest Singular Value/Eigenvalue of a Random Tensor

Spectral Theory 2021-06-22 v3 Optimization and Control Probability

Abstract

This short note presents upper bounds of the expectations of the largest singular values/eigenvalues of various types of random tensors in the non-asymptotic sense. For a standard Gaussian tensor of size n1××ndn_1\times\cdots\times n_d, it is shown that the expectation of its largest singular value is upper bounded by n1++nd\sqrt {n_1}+\cdots+\sqrt {n_d}. For the expectation of the largest d\ell^d-singular value, it is upper bounded by 2d12j=1dnjd22dj=1dnj122^{\frac{d-1}{2}}\prod_{j=1}^{d}n_j^{\frac{d-2}{2d}}\sum^d_{j=1}n_j^{\frac{1}{2}}. We also derive the upper bounds of the expectations of the largest Z-/H-(d\ell^d)/M-/C-eigenvalues of symmetric, partially symmetric, and piezoelectric-type Gaussian tensors, which are respectively upper bounded by dnd\sqrt n, d2d12nd12d\cdot 2^{\frac{d-1}{2}}n^{\frac{d-1}{2}}, 2m+2n2\sqrt m+2\sqrt n, and 3n3\sqrt n.

Keywords

Cite

@article{arxiv.2106.07433,
  title  = {On the Largest Singular Value/Eigenvalue of a Random Tensor},
  author = {Yuning Yang},
  journal= {arXiv preprint arXiv:2106.07433},
  year   = {2021}
}
R2 v1 2026-06-24T03:10:37.646Z