Solving quadratically constrained least squares using black box solvers (Q1198979)

The paper presents algorithms for solving quadratically constrained linear least squares problems that do not necessarily require expensive dense matrix factorizations. It is shown that only calls to a ``black box'' unconstrained solver are required. The authors' approach exploits the matrix structure inherent in the unconstrained problem. A formula for estimating the Lagrange multiplier which depends on the amount the unconstrained solution violates the constraint and an estimate of the smallest generalized singular value are derived. Numerical tests comparing the different algorithms are presented.