Abstract:An equivalent definition of hypermatrices is introduced. The matrix expression of hypermatrices is proposed. Using permutation matrices, the conversion of different matrix expressions is revealed. The various contracted products of hypermatrices are realized by semi-tensor products (STP) of matrices via matrix expressions of hypermatrices.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: **How to realize various contraction products of hypermatrices through the semi - tensor product (STP) of matrices, and reveal that STP is essentially a manifestation of multilinear mapping on hypermatrices**.
Specifically, the main objectives of the paper include:
1. **Introduce the equivalent definition of hypermatrices and their matrix representations**:
- Propose a new matrix expression method for hypermatrices.
- Use permutation matrices to show the conversion between different matrix expressions.
2. **Realize the contraction product of hypermatrices**:
- Through the matrix expressions of hypermatrices, use the semi - tensor product (STP) of matrices to realize various contraction products of hypermatrices.
- Prove that STP can be used to perform the contraction product of hypermatrices, thus revealing that STP is essentially a multilinear operator on hypermatrices.
3. **Explore the application of STP on hypermatrices**:
- Show that the action of STP on hypermatrices can be regarded as a generalization of the action of traditional matrix multiplication on matrices.
- Through specific examples and applications, illustrate the powerful functions of STP in dealing with hypermatrices.
### Formula Summary
- **Matrix expression of hypermatrices**:
\[
M_{r_1 \times r_2}^A \in F^{s \times t}, \quad s=\prod_{k = 1}^p n_{i_k}, \quad t=\prod_{\ell = 1}^q n_{j_\ell}
\]
- **Contraction product of hypermatrices**:
\[
C = A\times_{i_d}^{j_1}B
\]
where \(C\) is a hypermatrix, and its elements are:
\[
c_{i_1,\ldots,i_{d - 1},j_2,\ldots,j_r}=\sum_{k = 1}^n a_{i_1,\ldots,i_{d - 1},k}b_{k,j_2,\ldots,j_r}
\]
- **Contraction product under matrix expression**:
\[
M_{(i_1,\ldots,i_{d - 1})\times(j_2,\ldots,j_r)}^C = M_{(i_1,\ldots,i_{d - 1})\times i_d}^A\times M_{j_1\times(j_2,\ldots,j_r)}^B
\]
- **Definition of STP**:
\[
A\ltimes B:=(A\otimes I_{t/n})(B\otimes I_{t/p})
\]
Through these formulas and methods, the paper successfully transforms the complex operations of hypermatrices into the semi - tensor product operations of matrices, thus simplifying the calculations and revealing the essence of STP.