Upper bounds for singular values (Q1779404): Difference between revisions

Let \(A\) be an \(n\times n\) matrix, \(\sigma_1\geq \cdots \geq \sigma_n\) be the singular values of \(A\). It is known that \(\sigma_r=\min\| H\| \), where \(\| H\| \) is the operator norm of a matrix \(H\) with rank\((A+H)\leq r+1\). In this paper, the upper bounds for \(\sigma_r\) are obtained by various choices of \(H\).