Rank-Deficient Nonlinear Least Squares Problems and Subset Selection
نویسندگان
چکیده
We examine the local convergence of the Levenberg–Marquardt method for the solution of nonlinear least squares problems that are rank-deficient and have nonzero residual. We show that replacing the Jacobian by a truncated singular value decomposition can be numerically unstable. We recommend instead the use of subset selection. We corroborate our recommendations by perturbation analyses and numerical experiments.
منابع مشابه
Rank-deficient and Ill-conditioned Nonlinear Least Squares Problems
In this paper we review classical results by Ben-Israel, Boggs, and Tanabe on algorithms for rank-deficient nonlinear least squares problems and compare them to our recent perturbation theory results for such problems. Using a simple example, we show how sufficiently large perturbations can lead to poor convergence or even incorrect results, and how the application of subset selection can solve...
متن کاملOn the Evaluation Complexity of Cubic Regularization Methods for Potentially Rank-Deficient Nonlinear Least-Squares Problems and Its Relevance to Constrained Nonlinear Optimization
We propose a new termination criteria suitable for potentially singular, zero or non-zero residual, least-squares problems, with which cubic regularization variants take at most O(ǫ) residualand Jacobian-evaluations to drive either the Euclidean norm of the residual or its gradient below ǫ; this is the best-known bound for potentially rank-deficient nonlinear least-squares problems. We then app...
متن کاملSolving Rank-Deficient Linear Least-Squares Problems*
Numerical solution of linear least-squares problems is a key computational task in science and engineering. Effective algorithms have been developed for the linear least-squares problems in which the underlying matrices have full rank and are well-conditioned. However, there are few efficient and robust approaches to solving the linear least-squares problems in which the underlying matrices are...
متن کاملLocal results for the Gauss-Newton method on constrained rank-deficient nonlinear least squares
A nonlinear least squares problem with nonlinear constraints may be ill posed or even rank-deficient in two ways. Considering the problem formulated as minx 1/2‖f2(x)‖2 subject to the constraints f1(x) = 0, the Jacobian J1 = ∂f1/∂x and/or the Jacobian J = ∂f/∂x, f = [f1; f2], may be ill conditioned at the solution. We analyze the important special case when J1 and/or J do not have full rank at ...
متن کاملSuperlinearly convergent exact penalty projected structured Hessian updating schemes for constrained nonlinear least squares: asymptotic analysis
We present a structured algorithm for solving constrained nonlinear least squares problems, and establish its local two-step Q-superlinear convergence. The approach is based on an adaptive structured scheme due to Mahdavi-Amiri and Bartels of the exact penalty method of Coleman and Conn for nonlinearly constrained optimization problems. The structured adaptation also makes use of the ideas of N...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
- SIAM J. Numerical Analysis
دوره 49 شماره
صفحات -
تاریخ انتشار 2011