Scaling techniques for -subgradient methods
From MaRDI portal
Publication:2817840
DOI10.1137/14097642XzbMATH Open1347.65106arXiv1407.6133OpenAlexW2510723730MaRDI QIDQ2817840FDOQ2817840
Authors: S. Bonettini, V. Ruggiero, A. Benfenati 
Publication date: 2 September 2016
Published in: SIAM Journal on Optimization (Search for Journal in Brave)
Abstract: The recent literature on first order methods for smooth optimization shows that significant improvements on the practical convergence behaviour can be achieved with variable stepsize and scaling for the gradient, making this class of algorithms attractive for a variety of relevant applications. In this paper we introduce a variable metric in the context of the -subgradient projection methods for nonsmooth, constrained, convex problems, in combination with two different stepsize selection strategies. We develop the theoretical convergence analysis of the proposed approach and we also discuss practical implementation issues, as the choice of the scaling matrix. In order to illustrate the effectiveness of the method, we consider a specific problem in the image restoration framework and we numerically evaluate the effects of a variable scaling and of the steplength selection strategy on the convergence behaviour.
Full work available at URL: https://arxiv.org/abs/1407.6133
Recommendations
- Variable metric techniques for forward-backward methods in imaging
- \(\epsilon\)-subgradient algorithms for bilevel convex optimization
- Incremental subgradients for constrained convex optimization: A unified framework and new methods
- Variable metric inexact line-search-based methods for nonsmooth optimization
- New convergence results for the scaled gradient projection method
convergenceimage restorationvariable metricnonsmooth convex problemsforward-backward \(\epsilon\)-subgradient methodscaled primal-dual hybrid gradient algorithmstep size selection rulesTV restoration
Numerical mathematical programming methods (65K05) Convex programming (90C25) Image processing (compression, reconstruction, etc.) in information and communication theory (94A08) Numerical aspects of computer graphics, image analysis, and computational geometry (65D18)
Cites Work
- Nonlinear total variation based noise removal algorithms
- Convex analysis and monotone operator theory in Hilbert spaces
- Introductory lectures on convex optimization. A basic course.
- Convex optimization theory.
- Two-Point Step Size Gradient Methods
- Title not available (Why is that?)
- Convex Analysis
- On convergence rates of subgradient optimization methods
- Convergence of some algorithms for convex minimization
- Linear convergence of epsilon-subgradient descent methods for a class of convex functions
- Accelerated and inexact forward-backward algorithms
- A general framework for a class of first order primal-dual algorithms for convex optimization in imaging science
- A scaled gradient projection method for constrained image deblurring
- Image deblurring with Poisson data: from cells to galaxies
- Convergence of Approximate and Incremental Subgradient Methods for Convex Optimization
- Restoration of Poissonian Images Using Alternating Direction Optimization
- The cyclic Barzilai-–Borwein method for unconstrained optimization
- A descent proximal level bundle method for convex nondifferentiable optimization
- Title not available (Why is that?)
- Online learning and online convex optimization
- A primal-dual splitting method for convex optimization involving Lipschitzian, proximable and linear composite terms
- On the convergence of primal-dual hybrid gradient algorithm
- New adaptive stepsize selections in gradient methods
- Quasi-Fejérian analysis of some optimization algorithms
- On spectral properties of steepest descent methods
- Variable metric forward-backward splitting with applications to monotone inclusions in duality
- Variable metric quasi-Fejér monotonicity
- A limited memory steepest descent method
- An affine-scaling interior-point CBB method for box-constrained optimization
- Variable metric forward-backward algorithm for minimizing the sum of a differentiable function and a convex function
- Incremental subgradient methods for nondifferentiable optimization
- On the projected subgradient method for nonsmooth convex optimization in a Hilbert space
- Convergence of a simple subgradient level method
- An alternating extragradient method for total variation-based image restoration from Poisson data
- stochastic quasigradient methods and their application to system optimization†
- Error stability properties of generalized gradient-type algorithms
- Bregman operator splitting with variable stepsize for total variation image reconstruction
- Covariance-Preconditioned Iterative Methods for Nonnegatively Constrained Astronomical Imaging
- An inertial forward-backward algorithm for monotone inclusions
- Total variation-penalized Poisson likelihood estimation for ill-posed problems
- Convergence analysis of deflected conditional approximate subgradient methods
- Linear convergence of iterative soft-thresholding
- Non-negatively constrained image deblurring with an inexact interior point method
- Penalized maximum likelihood image restoration with positivity constraints: multiplicative algorithms
- The Efficiency of Subgradient Projection Methods for Convex Optimization, Part I: General Level Methods
- The Efficiency of Subgradient Projection Methods for Convex Optimization, Part II: Implementations and Extensions
- Projected subgradient methods with non-Euclidean distances for non-differentiable convex minimization and variational inequalities
- On the convergence of primal-dual hybrid gradient algorithms for total variation image restoration
- Scaling techniques for gradient projection-type methods in astronomical image deblurring
- Efficient gradient projection methods for edge-preserving removal of Poisson noise
- A scaled gradient projection method for Bayesian learning in dynamical systems
- On the convergence of conditional \(\varepsilon\)-subgradient methods for convex programs and convex-concave saddle-point problems.
- An \(\mathcal O(1/{k})\) convergence rate for the variable stepsize Bregman operator splitting algorithm
- A general method to devise maximum-likelihood signal restoration multiplicative algorithms with non-negativity constraints.
- Variable metric inexact line-search-based methods for nonsmooth optimization
- Nonnegative image reconstruction from sparse Fourier data: a new deconvolution algorithm
- A convergent blind deconvolution method for post-adaptive-optics astronomical imaging
- On proximal subgradient splitting method for minimizing the sum of two nonsmooth convex functions
- On the convergence of the forward-backward splitting method with linesearches
Cited In (7)
- \(\epsilon\)-subgradient algorithms for bilevel convex optimization
- A two-metric variable scaled forward-backward algorithm for \(\ell_0\) optimization problem and its applications
- Stochastic primal-dual hybrid gradient algorithm with adaptive step sizes
- Variable metric techniques for forward-backward methods in imaging
- Efficient position estimation of 3D fluorescent spherical beads in confocal microscopy via Poisson denoising
- Modified Fejér sequences and applications
- A Subspace Decomposition Principle for Scaled Gradient Projection Methods: Local Theory
This page was built for publication: Scaling techniques for \(\epsilon\)-subgradient methods
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2817840)
