An adaptive strategy for updating the damping parameters in an iteratively regularized Gauss-Newton method

In this paper a convergence analysis of an adaptive choice of the sequence of damping parameters in the iteratively regularized Gauss-Newton technique for solving nonlinear ill-posed operator equations is proposed. The selection criterion is motivated from parameter choice criteria, which are used for the efficient solution of nonlinear least-squares problems. The performance of this selection criterion is tested at nonlinear ill-posed model problems.