An adaptive cubic regularization method for computing extreme eigenvalues of tensors

TL;DR

This paper introduces an adaptive cubic regularization algorithm designed to efficiently compute the extreme H- and Z-eigenvalues of even order symmetric tensors, addressing a key challenge in tensor analysis.

Contribution

The paper presents a novel adaptive cubic regularization method specifically tailored for finding extreme eigenvalues of symmetric tensors, improving computational efficiency.

Findings

Successfully computes extreme eigenvalues of symmetric tensors

Demonstrates improved convergence over existing methods

Applicable to high-order tensor problems

Abstract

In this paper, we compute the H- and Z-eigenvalues of even order symmetric tensors by using the adaptive cubic regularization algorithm.