Geometrical inverse matrix approximation for least-squares problems and acceleration strategies

DOI10.1007/S11075-019-00862-ZzbMATH Open1456.65027arXiv1902.08388OpenAlexW3000692571MaRDI QIDQ827072FDOQ827072

M. Raydan, Jean-Paul Chehab

Publication date: 6 January 2021

Published in: Numerical Algorithms (Search for Journal in Brave)

Abstract: We extend the geometrical inverse approximation approach for solving linear least-squares problems. For that we focus on the minimization of $1-cos(X(A^{T}A),I)$, where $A$ is a given rectangular coefficient matrix and $X$ is the approximate inverse. In particular, we adapt the recently published simplified gradient-type iterative scheme MinCos to the least-squares scenario. In addition, we combine the generated convergent sequence of matrices with well-known acceleration strategies based on recently developed matrix extrapolation methods, and also with some deterministic and heuristic acceleration schemes which are based on affecting, in a convenient way, the steplength at each iteration. A set of numerical experiments, including large-scale problems, are presented to illustrate the performance of the different accelerations strategies.

Full work available at URL: https://arxiv.org/abs/1902.08388

Recommendations

• Generalized approximate inverse preconditioners for least squares problems
• A geometric Gauss-Newton method for least squares inverse eigenvalue problems
• Sur le problème des moindres carrés
• The least-squares solution of inverse problem over special matrices and its optimal approximation
• Pseudoinverse preconditioners and iterative methods for large dense linear least-squares problems

zbMATH Keywords

gradient-type methodsinverse approximationmatrix acceleration techniques

Mathematics Subject Classification ID

Ill-posedness and regularization problems in numerical linear algebra (65F22) Inverse problems in linear algebra (15A29)

Cites Work

• Damped Anderson Acceleration With Restarts and Monotonicity Control for Accelerating EM and EM-like Algorithms
• Anderson acceleration of the alternating projections method for computing the nearest correlation matrix
• Anderson Acceleration for Fixed-Point Iterations
• Iterative Procedures for Nonlinear Integral Equations
• New adaptive stepsize selections in gradient methods
• On spectral properties of steepest descent methods
• Extrapolation methods theory and practice
• Vector extrapolation methods. Applications and numerical comparison
• Matrix polynomial and epsilon-type extrapolation methods with applications
• Block extrapolation methods with applications
• The Simplified Topological $\varepsilon$-Algorithms for Accelerating Sequences in a Vector Space
• Gradient methods with adaptive step-sizes
• Approximate Inverse Techniques for Block-Partitioned Matrices
• Convergence acceleration during the 20th century
• An Element-Based Spectrally Optimized Approximate Inverse Preconditioner for the Euler Equations
• The epsilon algorithm and related topics
• Matrix differential equations and inverse preconditioners
• On the steplength selection in gradient methods for unconstrained optimization
• Preconditioned conjugate gradient method for finding minimal energy surfaces on Powell-Sabin triangulations
• A derivative-free nonmonotone line-search technique for unconstrained optimization
• Approximate inverse preconditioners for some large dense random electrostatic interaction matrices
• An analysis of sparse approximate inverse preconditioners for boundary integral equations
• Geometrical properties of the Frobenius condition number for positive definite matrices
• On the geometrical structure of symmetric matrices
• Optimal regularized low rank inverse approximation
• Sparse Approximate-Inverse Preconditioners Using Norm-Minimization Techniques
• Geometrical inverse preconditioning for symmetric positive definite matrices
• The simplified topological \(\varepsilon\)-algorithms: software and applications
• Approximate inverse computation using Frobenius inner product
• Résultats négatifs en accélération de la convergence
• Rational approximation to the Fermi-Dirac function with applications in density functional theory
• Sparse approximations of matrix functions via numerical integration of ODEs
• Convergence and acceleration properties for the vector \(\epsilon\)- algorithm
• The genesis and early developments of Aitken's process, Shanks' transformation, the \(\varepsilon\)-algorithm, and related fixed point methods
• Shanks Sequence Transformations and Anderson Acceleration
• Generalized approximate inverse preconditioners for least squares problems
• The Riemannian Barzilai–Borwein method with nonmonotone line search and the matrix geometric mean computation
• Enriched methods for large-scale unconstrained optimization

Cited In (4)

• A geometric theory for preconditioned inverse iteration. I: Extrema of Rayleigh quotient
• A new geometric acceleration of the von Neumann-Halperin projection method
• A geometric Gauss-Newton method for least squares inverse eigenvalue problems
• An accelerated tensorial double proximal gradient method for total variation regularization problem

Uses Software

• Anderson
• EPSfun

This page was built for publication: Geometrical inverse matrix approximation for least-squares problems and acceleration strategies

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q827072)