Shifted and extrapolated power methods for tensor ℓ^p-eigenpairs

09/26/2019
by   Stefano Cipolla, et al.
0

This work is concerned with the computation of ℓ^p-eigenvalues and eigenvectors of square tensors with d modes. In the first part we propose two possible shifted variants of the popular (higher-order) power method computation of ℓ^p-eigenpairs proving the convergence of both the schemes to the Perron ℓ^p-eigenvector of the tensor, and the maximal corresponding ℓ^p-eigenvalue, when the tensor is entrywise nonnegative and p is strictly larger than the number of modes. Then, motivated by the slow rate of convergence that the proposed methods achieve for certain real-world tensors, when p≈ d, the number of modes, in the second part we introduce an extrapolation framework based on the simplified topological ε-algorithm to efficiently accelerate the shifted power sequences. Numerical results on synthetic and real world problems show the improvements gained by the introduction of the shifting parameter and the efficiency of the acceleration technique.

READ FULL TEXT

page 1

page 2

page 3

page 4

03/07/2015

Higher order Matching Pursuit for Low Rank Tensor Learning

Low rank tensor learning, such as tensor completion and multilinear mult...
04/22/2022

Scalable symmetric Tucker tensor decomposition

We study the best low-rank Tucker decomposition of symmetric tensors, ad...
08/21/2021

Perturbation analysis of third-order tensor eigenvalue problem based on tensor-tensor multiplication

Perturbation analysis has been primarily considered to be one of the mai...
01/03/2022

Sketch-and-project methods for tensor linear systems

We first extend the famous sketch-and-project method and its adaptive va...
07/10/2019

Higher-order ergodicity coefficients for stochastic tensors

Ergodicity coefficients for stochastic matrices provide valuable upper b...
07/10/2019

Higher-order ergodicity coefficients

Ergodicity coefficients for stochastic matrices provide valuable upper b...
11/21/2017

Kullback-Leibler Principal Component for Tensors is not NP-hard

We study the problem of nonnegative rank-one approximation of a nonnegat...