TAOBRGN#

Bounded Regularized Gauss-Newton method for solving nonlinear least-squares problems with bound constraints. This algorithm is a thin wrapper around TAOBNTL that constructs the Gauss-Newton problem with the user-provided least-squares residual and Jacobian. The algorithm offers an L2-norm (“l2pure”), L2-norm proximal point (“l2prox”) regularizer, and L1-norm dictionary regularizer (“l1dict”), where we approximate the L1-norm ||x||_1 by sum_i(sqrt(x_i^2+epsilon^2)-epsilon) with a small positive number epsilon. Also offered is the “lm” regularizer which uses a scaled diagonal of J^T J. With the “lm” regularizer, TAOBRGN is a Levenberg-Marquardt optimizer. The user can also provide own regularization function.

Options Database Keys#

  • -tao_brgn_regularization_type - regularization type (“user”, “l2prox”, “l2pure”, “l1dict”, “lm”) (default “l2prox”)

  • -tao_brgn_regularizer_weight - regularizer weight (default 1e-4)

  • -tao_brgn_l1_smooth_epsilon - L1-norm smooth approximation parameter: ||x||_1 = sum(sqrt(x.^2+epsilon^2)-epsilon) (default 1e-6)

See Also#

Tao, TaoBRGNGetSubsolver(), TaoBRGNSetRegularizerWeight(), TaoBRGNSetL1SmoothEpsilon(), TaoBRGNSetDictionaryMatrix(), TaoBRGNSetRegularizerObjectiveAndGradientRoutine(), TaoBRGNSetRegularizerHessianRoutine()

Level#

beginner

Location#

src/tao/leastsquares/impls/brgn/brgn.c

Examples#

src/tao/leastsquares/tutorials/cs1.c
src/tao/leastsquares/tutorials/tomography.c


Edit on GitLab

Index of all Tao routines
Table of Contents for all manual pages
Index of all manual pages