Nonsmooth Regularization
نویسنده
چکیده
The purpose of this paper is to present to the Inverse Problems community some recent results regarding \nonsmooth", \edge preserving" regularization methods. Emphasis is placed on the penalized least squares (Tikhonov) regularization approach with penalty functionals of total variation type, although other nonquadratic penalty functionals are brieey discussed.
منابع مشابه
A Descent Method for Nonsmooth Variational Inequalities via Regularization
in this paper we propose a descent method for solving variational inequality problems where the underlying operator is nonsmooth, locally Lipschitz, and monotone over a closed, convex feasible set. The idea is to combine a descent method for variational inequality problems whose operators are nonsmooth, locally Lipschitz, and strongly monotone, with the Tikonov-Browder regularization technique....
متن کاملA Regularization Parameter for Nonsmooth Tikhonov Regularization
In this paper we develop a novel criterion for choosing regularization parameters for nonsmooth Tikhonov functionals. The proposed criterion is solely based on the value function, and thus applicable to a broad range of functionals. It is analytically compared with the local minimum criterion, and a posteriori error estimates are derived. An efficient numerical algorithm for computing the minim...
متن کاملOn Second-order Properties of the Moreau-Yosida Regularization for Constrained Nonsmooth Convex Programs
In this paper, we attempt to investigate a class of constrained nonsmooth convex optimization problems, that is, piecewise C2 convex objectives with smooth convex inequality constraints. By using the Moreau-Yosida regularization, we convert these problems into unconstrained smooth convex programs. Then, we investigate the second-order properties of the Moreau-Yosida regularization η. By introdu...
متن کاملMathematical Programming Manuscript No. Smooth Exact Penalty and Barrier Functions for Nonsmooth Optimization
For constrained nonsmooth optimization problems, continuously diierentiable penalty functions and barrier functions are given. They are proved exact in the sense that under some nondegeneracy assumption, local optimizers of a nonlinear program are also optimizers of the associated penalty or barrier function. This is achieved by augmenting the dimension of the program by a variable that control...
متن کاملA Two-Stage Method for Inverse Acoustic Medium Scattering
We present a novel numerical method to the time-harmonic inverse medium scattering problem of recovering the refractive index from noisy near-field scattered data. The approach consists of two stages, one pruning step of detecting the scatterer support, and one resolution enhancing step with nonsmooth mixed regularization. The first step is strictly direct and of sampling type, and it faithfull...
متن کامل