Backtracking strategies for accelerated descent methods with smooth composite objectives
From MaRDI portal
Abstract: We present and analyse a backtracking strategy for a general Fast Iterative Shrinkage/Thresholding Algorithm which has been recently proposed in (Chambolle, Pock, 2016) for strongly convex objective functions. Differently from classical Armijo-type line searching, our backtracking rule allows for local increasing and decreasing of the descent step size (i.e. proximal parameter) along the iterations. For such strategy accelerated convergence rates are proved and numerical results are shown for some exemplar imaging problems.
Recommendations
- Fast first-order methods for composite convex optimization with backtracking
- An acceleration of gradient descent algorithm with backtracking for unconstrained opti\-mi\-za\-tion
- Additive Schwarz methods for convex optimization with backtracking
- Gradient methods for minimizing composite functions
- Convex-concave backtracking for inertial Bregman proximal gradient algorithms in nonconvex optimization
Cites work
- (Nonlocal) total variation in medical imaging
- A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
- A Generalized Accelerated Composite Gradient Method: Uniting Nesterov's Fast Gradient Method and FISTA
- A remark on accelerated block coordinate descent for computing the proximity operators of a sum of convex functions
- A variable metric forward-backward method with extrapolation
- Accelerated and inexact forward-backward algorithms
- Adaptive restart for accelerated gradient schemes
- An Accelerated Composite Gradient Method for Large-Scale Composite Objective Problems
- An adaptive accelerated proximal gradient method and its homotopy continuation for sparse optimization
- An algorithm for total variation minimization and applications
- An introduction to continuous optimization for imaging
- Convex programming in Hilbert space
- Fast first-order methods for composite convex optimization with backtracking
- Fast Gradient-Based Algorithms for Constrained Total Variation Image Denoising and Deblurring Problems
- First order algorithms in variational image processing
- Gradient methods for minimizing composite functions
- scientific article; zbMATH DE number 3850830 (Why is no real title available?)
- Inexact and accelerated proximal point algorithms
- Introductory lectures on convex optimization. A basic course.
- Local Linear Convergence of ISTA and FISTA on the LASSO Problem
- Minimization of functions having Lipschitz continuous first partial derivatives
- New Proximal Point Algorithms for Convex Minimization
- Nonlinear total variation based noise removal algorithms
- On the convergence of the iterates of the ``fast iterative shrinkage/thresholding algorithm
- Regularization and Variable Selection Via the Elastic Net
- Signal Recovery by Proximal Forward-Backward Splitting
- Smooth minimization of non-smooth functions
- Stability of over-relaxations for the forward-backward algorithm, application to FISTA
- Stochastic Primal-Dual Hybrid Gradient Algorithm with Arbitrary Sampling and Imaging Applications
- The doubly regularized support vector machine
Cited in
(30)- Convergence rates of an inertial gradient descent algorithm under growth and flatness conditions
- Additive Schwarz methods for convex optimization with backtracking
- Convergence rate of inertial proximal algorithms with general extrapolation and proximal coefficients
- Accelerated variational PDEs for efficient solution of regularized inversion problems
- Fast gradient methods for uniformly convex and weakly smooth problems
- An abstract convergence framework with application to inertial inexact forward-backward methods
- Fast first-order methods for composite convex optimization with backtracking
- An optimal high-order tensor method for convex optimization
- Convex-concave backtracking for inertial Bregman proximal gradient algorithms in nonconvex optimization
- Improving ``fast iterative shrinkage-thresholding algorithm: faster, smarter, and greedier
- A unified adaptive tensor approximation scheme to accelerate composite convex optimization
- Compressive Computed Tomography Reconstruction through Denoising Approximate Message Passing
- Scaled, inexact, and adaptive generalized FISTA for strongly convex optimization
- An Optimized Dynamic Mode Decomposition Model Robust to Multiplicative Noise
- Linearly-convergent FISTA variant for composite optimization with duality
- A dual-primal finite element tearing and interconnecting method for nonlinear variational inequalities utilizing linear local problems
- Parameter-free accelerated gradient descent for nonconvex minimization
- Parameter-free FISTA by adaptive restart and backtracking
- Additive Schwarz methods for convex optimization -- convergence theory and acceleration
- Additive Schwarz methods for fourth-order variational inequalities
- Several kinds of acceleration techniques for unconstrained optimization first-order algorithms
- A C-FISTA-type Krasnosel'skiĭ-Mann iteration with applications to TV-based denoising problem and image in-painting
- A fast algorithm for smooth convex minimization problems and its application to inverse source problems
- Whiteness-based bilevel estimation of weighted TV parameter maps for image denoising
- Efficient parameter-free restarted accelerated gradient methods for convex and strongly convex optimization
- FISTA restart using an automatic estimation of the growth parameter
- Improved accelerated gradient algorithms with line search for smooth convex optimization problems
- C-FISTA type projection algorithm for quasi-variational inequalities
- A C-FISTA-type proximal point algorithm for strongly quasiconvex pseudomonotone equilibrium problems
- A scaled and adaptive FISTA algorithm for signal-dependent sparse image super-resolution problems
This page was built for publication: Backtracking strategies for accelerated descent methods with smooth composite objectives
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5231684)
