An integral representation of the determinant of a matrix and its applications
Joshua Du, Jun Hyeok Ji
Abstract
Joshua Du, Jun Hyeok Ji
Abstract
The Hadamard determinant theorem states that the ratio of the determinant of a square matrix over the complex field to the product of its main diagonal elements is less than or equal to one for a positive definite Hermitian matrix. An integral representation of this ratio for both positive definite Hermitian matrix and diagonally dominant real matrix is given in this paper. Using this new identity, an alternative proof of the famous Hadamard determinant theorem is discussed. In addition, a lower bound of determinant in terms of the product of the main diagonal elements is given. Finally, a numerical algorithm fundamentally different from current approaches in literature is also proposed for the computation of the determinant of a small and dense matrix. Numerical experiments indicate that this new approach is robust.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The Hadamard determinant theorem states that the ratio of the determinant of a square matrix over the complex field to the product of its main diagonal elements is less than or equal to one for a positive definite Hermitian matrix. An integral representation of this ratio for both positive definite Hermitian matrix and diagonally dominant real matrix is given in this paper. Using this new identity, an alternative proof of the famous Hadamard determinant theorem is discussed. In addition, a lower bound of determinant in terms of the product of the main diagonal elements is given. Finally, a numerical algorithm fundamentally different from current approaches in literature is also proposed for the computation of the determinant of a small and dense matrix. Numerical experiments indicate that this new approach is robust.
Key concepts: Mathematics, Hermitian matrix, Square matrix, Block matrix, Matrix (chemical analysis), Hadamard product, Identity matrix, Hadamard transform