Nonmonotone globalization for Anderson acceleration via adaptive regularization
From MaRDI portal
Publication:6159247
Abstract: Anderson acceleration (AA) is a popular method for accelerating fixed-point iterations, but may suffer from instability and stagnation. We propose a globalization method for AA to improve stability and achieve unified global and local convergence. Unlike existing AA globalization approaches that rely on safeguarding operations and might hinder fast local convergence, we adopt a nonmonotone trust-region framework and introduce an adaptive quadratic regularization together with a tailored acceptance mechanism. We prove global convergence and show that our algorithm attains the same local convergence as AA under appropriate assumptions. The effectiveness of our method is demonstrated in several numerical experiments.
Recommendations
- A Proof That Anderson Acceleration Improves the Convergence Rate in Linearly Converging Fixed-Point Methods (But Not in Those Converging Quadratically)
- Convergence analysis for Anderson acceleration
- Anderson acceleration for fixed-point iterations
- Anderson Acceleration for Nonsmooth Fixed Point Problems
- Anderson acceleration based on the \(\mathcal{H}^{- s}\) Sobolev norm for contractive and noncontractive fixed-point operators
Cites work
- A characterization of the behavior of the Anderson acceleration on linear problems
- A comparative study on methods for convergence acceleration of iterative vector sequences
- A method for the solution of certain non-linear problems in least squares
- A New Alternating Minimization Algorithm for Total Variation Image Reconstruction
- A nonlinear GMRES optimization algorithm for canonical tensor decomposition
- A nonsmooth version of Newton's method
- A Proof That Anderson Acceleration Improves the Convergence Rate in Linearly Converging Fixed-Point Methods (But Not in Those Converging Quadratically)
- Activity identification and local linear convergence of forward-backward-type methods
- An Algorithm for Least-Squares Estimation of Nonlinear Parameters
- An analysis for the DIIS acceleration method used in quantum chemistry calculations
- An elementary proof of convergence for the forward-backward splitting algorithm
- Anderson Accelerated Douglas--Rachford Splitting
- Anderson accelerated fixed-stress splitting schemes for consolidation of unsaturated porous media
- Anderson acceleration and application to the three-temperature energy equations
- Anderson acceleration for a class of nonsmooth fixed-point problems
- Anderson acceleration for contractive and noncontractive operators
- Anderson acceleration for fixed-point iterations
- Anderson acceleration method of finding steady-state particle size distribution for a wide class of aggregation-fragmentation models
- Anderson acceleration of the Jacobi iterative method: an efficient alternative to Krylov methods for large, sparse linear systems
- Anderson-accelerated convergence of Picard iterations for incompressible Navier-Stokes equations
- Comments on: ``Anderson acceleration, mixing and extrapolation
- Convergence analysis for Anderson acceleration
- Convex analysis and monotone operator theory in Hilbert spaces
- Forward-backward quasi-Newton methods for nonsmooth optimization problems
- Generalized Hessian Properties of Regularized Nonsmooth Functions
- Globally Convergent Type-I Anderson Acceleration for Nonsmooth Fixed-Point Iterations
- scientific article; zbMATH DE number 46303 (Why is no real title available?)
- scientific article; zbMATH DE number 1998925 (Why is no real title available?)
- scientific article; zbMATH DE number 5937962 (Why is no real title available?)
- Implicit Functions and Solution Mappings
- Introduction to nonlinear optimization: theory, algorithms, and applications with MATLAB
- Iterative Procedures for Nonlinear Integral Equations
- Linear Convergence and Metric Selection for Douglas-Rachford Splitting and ADMM
- Local convergence properties of Douglas-Rachford and alternating direction method of multipliers
- Local improvement results for Anderson acceleration with inaccurate function evaluations
- Nonmonotone trust-region methods for bound-constrained semismooth equations with applications to nonlinear mixed complementarity problems
- On the asymptotic linear convergence speed of Anderson acceleration applied to ADMM
- Regularized nonlinear acceleration
- Spectral operators of matrices: semismoothness and characterizations of the generalized Jacobian
- Strong Semismoothness of Eigenvalues of Symmetric Matrices and Its Application to Inverse Eigenvalue Problems
- Strong semismoothness of the Fischer-Burmeister SDC and SOC complementarity functions
- Trust Region Methods
- Two classes of multisecant methods for nonlinear acceleration
Cited in
(9)- Anderson acceleration for degenerate and nondegenerate problems
- Globally Convergent Type-I Anderson Acceleration for Nonsmooth Fixed-Point Iterations
- Descent Properties of an Anderson Accelerated Gradient Method with Restarting
- Anderson acceleration with truncated Gram-Schmidt
- An extra gradient Anderson-accelerated algorithm for pseudomonotone variational inequalities
- Anderson acceleration of derivative-free projection methods for constrained monotone nonlinear equations
- A fast Landweber-type iterative regularization method based on Anderson acceleration for nonlinear ill-posed problems
- A nonmonotone globalization framework for Anderson acceleration for contractive and nonexpansive fixed point problems
- A smoothing Anderson acceleration algorithm for nonsmooth fixed point problems with linear convergence
This page was built for publication: Nonmonotone globalization for Anderson acceleration via adaptive regularization
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6159247)
