HOID: Higher Order Interpolatory Decomposition for tensors based on Tucker representation

TL;DR

This paper introduces HOID, a tensor decomposition method based on interpolatory techniques in Tucker format, which preserves important features like sparsity and non-negativity, with validated error bounds and computational analysis.

Contribution

The paper develops a novel Higher Order Interpolatory Decomposition (HOID) for tensors that extends matrix CUR methods to tensor data, preserving key features and providing error bounds.

Findings

HOID effectively preserves sparsity and non-negativity.

The method offers competitive error bounds and computational efficiency.

Numerical examples demonstrate the advantages of HOID over existing methods.

Abstract

We derive a CUR-type factorization for tensors in the Tucker format based on interpolatory decomposition, which we will denote as Higher Order Interpolatory Decomposition (HOID). Given a tensor $\mathcal{X}$, the algorithm provides a set of column vectors $\{ \mathbf{C}_n\}_{n=1}^d$ which are columns extracted from the mode-$n$ tensor unfolding, along with a core tensor $\mathcal{G}$ and together, they satisfy some error bounds. Compared to the Higher Order SVD (HOSVD) algorithm, the HOID provides a decomposition that preserves certain important features of the original tensor such as sparsity, non-negativity, integer values, etc. Error bounds along with detailed estimates of computational costs are provided. The algorithms proposed in this paper have been validated against carefully chosen numerical examples which highlight the favorable properties of the algorithms. Related methods for subset selection proposed for matrix CUR decomposition, such as Discrete Empirical Interpolation method (DEIM) and leverage score sampling, have also been extended to tensors and are compared against our proposed algorithms.