A Constructive Algorithm for Decomposing a Tensor into a Finite Sum of Orthonormal Rank-1 Terms

TL;DR

This paper introduces TTr1SVD, a novel constructive algorithm that decomposes real tensors into orthonormal rank-1 terms, generalizing SVD for tensors with unique properties and practical error quantification.

Contribution

The paper presents TTr1SVD, a new tensor decomposition method that guarantees orthonormality, uniqueness, and provides a complete characterization of orthogonal tensors, along with conversion to Tucker format.

Findings

TTr1SVD decomposes tensors into orthonormal rank-1 terms.

The method guarantees uniqueness for a fixed index order.

Numerical examples demonstrate the properties and effectiveness of the decomposition.

Abstract

We propose a constructive algorithm that decomposes an arbitrary real tensor into a finite sum of orthonormal rank-1 outer products. The algorithm, named TTr1SVD, works by converting the tensor into a tensor-train rank-1 (TTr1) series via the singular value decomposition (SVD). TTr1SVD naturally generalizes the SVD to the tensor regime with properties such as uniqueness for a fixed order of indices, orthogonal rank-1 outer product terms, and easy truncation error quantification. Using an outer product column table it also allows, for the first time, a complete characterization of all tensors orthogonal with the original tensor. Incidentally, this leads to a strikingly simple constructive proof showing that the maximum rank of a real $2 \times 2 \times 2$ tensor over the real field is 3. We also derive a conversion of the TTr1 decomposition into a Tucker decomposition with a sparse core tensor. Numerical examples illustrate each of the favorable properties of the TTr1 decomposition.