📝 SummarXiv

Abstract

This paper provides a general solution for the Kronecker product decomposition (KPD) of vectors, matrices, and hypermatrices. First, an algorithm, namely, monic decomposition algorithm (MDA), is reviewed. It consists of a set of projections from a higher dimension Euclidian space to its factor-dimension subspaces. It is then proved that the KPD of vectors is solvable, if and only if, the project mappings provide the required decomposed vectors. Hence it provides an easily verifiable necessary and sufficient condition for the KPD of vectors. Then an algorithm is proposed to calculate the least square error approximated decomposition. Using it finite times a finite sum (precise) KPD of any vectors can be obtained. Then the swap matrix is used to make the elements of a matrix re-arranging, and then provides a method to convert the KPD of matrices to its corresponding KPD of vectors. It is proved that the KPD of a matrix is solvable, if and only if, the KPD of its corresponding vector is solvable. In this way, the necessary and sufficient condition, and the followed algorithms for approximate and finite sum KPDs for matrices are also obtained. Finally, the permutation matrix is introduced and used to convert the KPD of any hypermatrix to KPD of its corresponding vector. Similarly to matrix case, the necessary and sufficient conditions for solvability and the techniques for vectors and matrices are also applicable for hypermatrices, though some additional algorithms are necessary. Several numerical examples are included to demonstrate this universal KPD solving method.