Activity Identification and Local Linear Convergence of Inertial Forward-Backward Splitting
Explore this paper's citation graph
Summary
A unified analysis is proposed, under which it is shown that iFB-type splitting correctly identifies the active manifold M in a finite number of iterations, and then enters a local (linear) convergence regime, which is characterised precisely.
- Type
- preprint
- Published
- 2015-03-12
- Cited by
- 9
- References
- 41
- Access
- Open access
- OpenAlex
- https://openalex.org/W189176795
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:118053310
Keywords
Lipschitz continuity, Inertial frame of reference, Mathematics, Manifold (fluid mechanics), Regularization (linguistics)
References
- ε-Enlargements of Maximal Monotone Operators in Banach Spaces
- Identifying active constraints via partial smoothness and prox-regularity
- Convex Analysis and Monotone Operator Theory in Hilbert Spaces
- An Inertial Forward-Backward Algorithm for Monotone Inclusions
- Fast Convergence of an Inertial Gradient-like System with Vanishing Viscosity
- Local Linear Convergence of ISTA and FISTA on the LASSO Problem
- Introduction to optimization
- Partly Smooth Regularization of Inverse Problems
- Sparse Spikes Deconvolution on Thin Grids
- The Rate of Convergence of Nesterov's Accelerated Forward-Backward Method is Actually Faster Than 1/k2
- Optimization Techniques on Riemannian Manifolds
- Local and Global Convergence of an Inertial Version of Forward-Backward Splitting
- Linear Convergence of Iterative Soft-Thresholding
- Convergence of a splitting inertial proximal method for monotone operators
- Fixed-Point Continuation for l1-Minimization: Methodology and Convergence
- Splitting Algorithms for the Sum of Two Nonlinear Operators
- The [barred L]ojasiewicz Inequality for Nonsmooth Subanalytic Functions with Applications to Subgradient Dynamical Systems
- Geometrical interpretation of the predictor-corrector type algorithms in structured optimization problems
- Monotone Operators and the Proximal Point Algorithm
- Weak convergence of the sequence of successive approximations for nonexpansive mappings
Cited by
- Real-time ℓ^1 -- ℓ^2 deblurring using wavelet expansions of operators
- Fast convex optimization via inertial dynamics with Hessian driven damping
- Local and global convergence of a general inertial proximal splitting scheme for minimizing composite functions
- Local Q-linear convergence and finite-time active set identification of ADMM on a class of penalized regression problems
- Local Convergence Properties of Douglas--Rachford and ADMM
- Convergence rates of inertial splitting schemes for nonconvex composite optimization
- Local Convergence Properties of Douglas–Rachford and Alternating Direction Method of Multipliers
- Enhancement of functional brain connectome analysis by the use of deformable models in the estimation of spatial decompositions of the brain images. (Amélioration de connectivité fonctionnelle par utilisation de modèles déformables dans l'estimation de décompositions spatiales des images de cerveau)
- A Forward–Backward Algorithm With Different Inertial Terms for Structured Non-Convex Minimization Problems
Related papers
- A relative-error inertial-relaxed inexact projective splitting algorithm
- Parallel random block-coordinate forward–backward algorithm: a unified convergence analysis
- Convergence Rates of Distributed Gradient Methods Under Random Quantization: A Stochastic Approximation Approach
- High-dimensional dynamic systems identification with additional constraints
- Properties of linear systems in PDE-constrained optimization. Part I: Distributed control 1
- A Characterization of Stochastic Mirror Descent Algorithms and Their Convergence Properties
- Sharp global convergence guarantees for iterative nonconvex optimization: A Gaussian process perspective
- A-priori error analysis of local incremental minimization schemes for rate-independent evolutions