A Bregman forward-backward linesearch algorithm for nonconvex composite optimization: superlinear convergence to nonisolated local minima
From MaRDI portal
(Redirected from Publication:5853567)
Abstract: We introduce Bella, a locally superlinearly convergent Bregman forward backward splitting method for minimizing the sum of two nonconvex functions, one of which satisfying a relative smoothness condition and the other one possibly nonsmooth. A key tool of our methodology is the Bregman forward-backward envelope (BFBE), an exact and continuous penalty function with favorable first- and second-order properties, and enjoying a nonlinear error bound when the objective function satisfies a Lojasiewicz-type property. The proposed algorithm is of linesearch type over the BFBE along candidate update directions, and converges subsequentially to stationary points, globally under a KL condition, and owing to the given nonlinear error bound can attain superlinear convergence rates even when the limit point is a nonisolated minimum, provided the directions are suitably selected.
Recommendations
- Forward-backward envelope for the sum of two nonconvex functions: further properties and nonmonotone linesearch algorithms
- Forward-backward quasi-Newton methods for nonsmooth optimization problems
- Non-smooth non-convex Bregman minimization: unification and new algorithms
- Inertial proximal gradient methods with Bregman regularization for a class of nonconvex optimization problems
- Bregman Finito/MISO for nonconvex regularized finite sum minimization without Lipschitz gradient continuity
Cites work
- A Characterization of Superlinear Convergence and Its Application to Quasi-Newton Methods
- A coordinate gradient descent method for nonsmooth separable minimization
- A descent lemma beyond Lipschitz gradient continuity: first-order methods revisited and applications
- A Fast Iterative Shrinkage-Thresholding Algorithm for Linear Inverse Problems
- A generalized proximal point algorithm for certain non-convex minimization problems
- A geometric analysis of phase retrieval
- A Globally and Superlinearly Convergent Algorithm for Nonsmooth Convex Minimization
- A highly efficient semismooth Newton augmented Lagrangian method for solving lasso problems
- A Proximal Minimization Algorithm for Structured Nonconvex and Nonsmooth Problems
- A regularized Newton method without line search for unconstrained optimization
- A simplified view of first order methods for optimization
- Accelerated first-order methods for large-scale convex optimization: nearly optimal complexity under strong convexity
- An envelope for Davis-Yin splitting and strict saddle-point avoidance
- An inertial forward-backward algorithm for the minimization of the sum of two nonconvex functions
- An inertial Tseng's type proximal algorithm for nonsmooth and nonconvex optimization problems
- Behavior of accelerated gradient methods near critical points of nonconvex functions
- Bregman proximal mappings and Bregman-Moreau envelopes under relative prox-regularity
- Characterizations of Łojasiewicz inequalities: Subgradient flows, talweg, convexity
- Clarke Subgradients of Stratifiable Functions
- Convergence of descent methods for semi-algebraic and tame problems: proximal algorithms, forward-backward splitting, and regularized Gauss-Seidel methods
- Convergence of non-smooth descent methods using the Kurdyka-Łojasiewicz inequality
- Convex Analysis
- Convex analysis and monotone operator theory in Hilbert spaces
- Convex-concave backtracking for inertial Bregman proximal gradient algorithms in nonconvex optimization
- Douglas-Rachford splitting for nonconvex optimization with application to nonconvex feasibility problems
- Entropic Proximal Mappings with Applications to Nonlinear Programming
- ESSENTIAL SMOOTHNESS, ESSENTIAL STRICT CONVEXITY, AND LEGENDRE FUNCTIONS IN BANACH SPACES
- Finding zeros of Hölder metrically subregular mappings via globally convergent Levenberg-Marquardt methods
- Finite-Dimensional Variational Inequalities and Complementarity Problems
- First order methods beyond convexity and Lipschitz gradient continuity with applications to quadratic inverse problems
- First-order methods almost always avoid strict saddle points
- Forward-backward envelope for the sum of two nonconvex functions: further properties and nonmonotone linesearch algorithms
- Forward-backward quasi-Newton methods for nonsmooth optimization problems
- Forward-backward splitting with Bregman distances
- From error bounds to the complexity of first-order descent methods for convex functions
- Further properties of the forward-backward envelope with applications to difference-of-convex programming
- Generalizations of the Dennis--Moré Theorem
- Generalized Hessian Properties of Regularized Nonsmooth Functions
- Geometric categories and o-minimal structures
- Gradient methods for minimizing composite functions
- scientific article; zbMATH DE number 1694914 (Why is no real title available?)
- scientific article; zbMATH DE number 5077050 (Why is no real title available?)
- scientific article; zbMATH DE number 1329071 (Why is no real title available?)
- scientific article; zbMATH DE number 1160037 (Why is no real title available?)
- scientific article; zbMATH DE number 5060482 (Why is no real title available?)
- scientific article; zbMATH DE number 3371284 (Why is no real title available?)
- Implementable tensor methods in unconstrained convex optimization
- Incremental majorization-minimization optimization with application to large-scale machine learning
- Local behavior of an iterative framework for generalized equations with nonisolated solutions
- Local convergence of the Levenberg-Marquardt method under Hölder metric subregularity
- Low rank matrix completion by alternating steepest descent methods
- Majorization-minimization procedures and convergence of SQP methods for semi-algebraic and tame programs
- Newton-Type Methods for Optimization and Variational Problems
- Non-smooth non-convex Bregman minimization: unification and new algorithms
- Nonlinear error bounds via a change of function
- On gradients of functions definable in o-minimal structures
- On linear convergence of non-Euclidean gradient methods without strong convexity and Lipschitz gradient continuity
- On semi- and subanalytic geometry
- On the quadratic convergence of the cubic regularization method under a local error bound condition
- Optimal subgradient methods: computational properties for large-scale linear inverse problems
- Practical Aspects of the Moreau--Yosida Regularization: Theoretical Preliminaries
- Properties of the Moreau-Yosida regularization of a piecewise C^2 convex function
- Prox-regular functions in variational analysis
- Proximal alternating linearized minimization for nonconvex and nonsmooth problems
- Proximal Alternating Minimization and Projection Methods for Nonconvex Problems: An Approach Based on the Kurdyka-Łojasiewicz Inequality
- Proximité et dualité dans un espace hilbertien
- Quasi-Newton Methods, Motivation and Theory
- Regularized Newton methods for convex minimization problems with singular solutions
- Regularizing with Bregman-Moreau envelopes
- Relatively smooth convex optimization by first-order methods, and applications
- SuperMann: A Superlinearly Convergent Algorithm for Finding Fixed Points of Nonexpansive Operators
- The Moreau envelope function and proximal mapping in the sense of the Bregman distance
- The Łojasiewicz Inequality for Nonsmooth Subanalytic Functions with Applications to Subgradient Dynamical Systems
- Variable metric inexact line-search-based methods for nonsmooth optimization
- Variational Analysis
Cited in
(32)- Multi-block Bregman proximal alternating linearized minimization and its application to orthogonal nonnegative matrix factorization
- A block inertial Bregman proximal algorithm for nonsmooth nonconvex problems with application to symmetric nonnegative matrix tri-factorization
- Proximal gradient algorithms under local Lipschitz gradient continuity. A convergence and robustness analysis of PANOC
- Accelerated first-order methods for large-scale convex optimization: nearly optimal complexity under strong convexity
- Forward-backward envelope for the sum of two nonconvex functions: further properties and nonmonotone linesearch algorithms
- Block Bregman majorization minimization with extrapolation
- Nonlinear forward-backward splitting with projection correction
- A stochastic semismooth Newton method for nonsmooth nonconvex optimization
- Bregman Finito/MISO for nonconvex regularized finite sum minimization without Lipschitz gradient continuity
- Dualities for Non-Euclidean Smoothness and Strong Convexity under the Light of Generalized Conjugacy
- A Bregman stochastic method for nonconvex nonsmooth problem beyond global Lipschitz gradient continuity
- Smoothing unadjusted Langevin algorithms for nonsmooth composite potential functions
- Stochastic composition optimization of functions without Lipschitz continuous gradient
- A Regularized Newton Method for \({\boldsymbol{\ell}}_{q}\) -Norm Composite Optimization Problems
- A class of modified accelerated proximal gradient methods for nonsmooth and nonconvex minimization problems
- An approximate Newton-type proximal method using symmetric rank-one updating formula for minimizing the nonsmooth composite functions
- SPIRAL: a superlinearly convergent incremental proximal algorithm for nonconvex finite sum minimization
- A scalar Hessian estimation with a sparse nonmonotone line search technique for the sparse recovery problem
- A Bregman proximal subgradient algorithm for nonconvex and nonsmooth fractional optimization problems
- An interior proximal gradient method for nonconvex optimization
- Extrapolated plug-and-play three-operator splitting methods for nonconvex optimization with applications to image restoration
- High-order methods beyond the classical complexity bounds: inexact high-order proximal-point methods
- A mirror inertial forward-reflected-backward splitting: convergence analysis beyond convexity and Lipschitz smoothness
- Variance-reduction for variational inequality problems with Bregman distance function
- Moreau envelope and proximal-point methods under the lens of high-order regularization
- Smoothing randomized block-coordinate proximal gradient algorithms for nonsmooth nonconvex composite optimization
- Quasi-Newton type proximal gradient method for nonconvex nonsmooth composite optimization problems
- On the forward-backward method with nonmonotone linesearch for infinite-dimensional nonsmooth nonconvex problems
- An inexact q-order regularized proximal Newton method for nonconvex composite optimization
- On a globally convergent semismooth^* Newton method in nonsmooth nonconvex optimization
- An inertial Bregman proximal gradient algorithm for nonconvex composite optimization problems
- An inexact projected regularized Newton method for fused zero-norms regularization problems
This page was built for publication: A Bregman forward-backward linesearch algorithm for nonconvex composite optimization: superlinear convergence to nonisolated local minima
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5853567)
