A logical implication between two conjectures on matrix permanents
Let \(H_n\) be the set of \(n \times n\) positive semidefinite Hermitian matrices and \(S_n\) be the symmetric group. The permanent of a \(n \times n\) matrix \(A\) is defined as \N\[\N\mathrm{per} (A) = \sum_{\sigma \in S_n} \prod_{i=1}^n a_{i,\sigma(i)}.\N\]\NFor a given matrix \(A\), let \(A(i,j)\) be the \((n-1) \times (n-1)\) submatrix of \(A\) obtained by deleting the \(i\)-th row and \(j\)-th column of \(A\).\N\N\textit{R. B. Bapat} and \textit{V. S. Sunder} [Linear Algebra Appl. 76, 153--163 (1986; Zbl 0602.15008)] conjectured the following two inequalities:\N\NConjecture 1. If \(A, B \in H_n\), then \N\[\N\mathrm{per} (A \circ B) \leq \mathrm{per} (A) \prod_{i=1}^n b_{i,i}, \tag{1} \N\]\Nwhere \(A \circ B\) denotes the Hadamard (or entry-wise) matrix product.\N\NConjecture 2. If \(A \in H_n\), then \N\[\N\lambda \leq\mathrm{per} (A), \tag{2}\N\]\Nwhere \(\lambda\) denotes the largest eigenvalue of the matrix \(F_A\), which is deduced from matrix \(A\) following \(f_{i,j} = a_{i,j} \mathrm{per} (A(i, j))\).\N\N\textit{S. W. Drury} [Electron. J. Linear Algebra 31, 69--70 (2016; Zbl 1332.15019); Math. Inequal. Appl. 21, No. 2, 517--520 (2018; Zbl 1384.15004)] exhibited a counterexample of the two aforementioned conjectures. However, in this paper, the authors make some study of both conjectures. The main result is the following:\NIf \(A \in H_n\) is chosen such that inequality (1) is verified for all \(B \in H_n\), then \(A\) must also verify inequality (2).\N\NThe authors finish the article showing some consequences of the main theorem.
- A counterexample to a question of Bapat \& Sunder
- A counterexample to a question of Bapat and Sunder
- A real counterexample to two inequalities involving permanents
- An analytic approach to a permanent conjecture
- An extremal property of the permanent and the determinant
- An update on a few permanent conjectures
- scientific article; zbMATH DE number 3232824 (Why is no real title available?)
- Inequalities connected with definite Hermitian forms.
- Is There a Permanental Analogue to Oppenheim's Inequality?
- Lieb's permanental dominance conjecture
- Notes on hadamard products of matrices
- On majorization and Schur products
- On permanental compounds
This page was built for publication: A logical implication between two conjectures on matrix permanents
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6932149)