Iterative methods to compute weighted normal pseudosolutions with positive definite weights (Q1280950): Difference between revisions

Two iterative methods to compute the weighted normal pseudosolution for a linear system with an arbitrary rectangular real matrix are constructed and proved that the convergence is linear. The first process requires finding the maximum and minimum nonzero eigenvalue of a symmetrizable positive semidefinite matrix. The second process requires inverting a symmetrizable positive matrix.