Singular value and generalized singular value decompositions and the solution of linear matrix equations (Q1087955): Difference between revisions

All matrices are taken to be real, but not necessarily square. The author employs the singular value decomposition and a generalization of it to study the solvability of the equation \(AXB+CYD=E\), and of the pair of equations \(AXB=E\), \(FXG=H\). A couple of special cases are considered, and numerical algorithms for the solutions are suggested.