Independent component analysis for tensor-valued data

TL;DR

This paper introduces tensor-specific ICA methods, MFOBI and TFOBI, which preserve tensor structure during analysis, offering improved performance over traditional vectorized approaches, demonstrated through simulations and real data clustering.

Contribution

The paper proposes tensor extensions of FOBI, MFOBI and TFOBI, with algorithms and asymptotic theory, to better analyze tensor-valued data without vectorization.

Findings

MFOBI and TFOBI outperform classical FOBI in simulations

The methods effectively recover independent components in tensor data

Application to real data demonstrates practical utility

Abstract

In preprocessing tensor-valued data, e.g. images and videos, a common procedure is to vectorize the observations and subject the resulting vectors to one of the many methods used for independent component analysis (ICA). However, the tensor structure of the original data is lost in the vectorization and, as a more suitable alternative, we propose the matrix- and tensor fourth order blind identification (MFOBI and TFOBI). In these tensorial extensions of the classic fourth order blind identification (FOBI) we assume a Kronecker structure for the mixing and perform FOBI simultaneously on each direction of the observed tensors. We discuss the theory and assumptions behind MFOBI and TFOBI and provide two different algorithms and related estimates of the unmixing matrices along with their asymptotic properties. Finally, simulations are used to compare the method's performance with that of classical FOBI for vectorized data and we end with a real data clustering example.