Abstract:Abstract Recently, circumcentering reflection method (CRM) has been introduced for solving the feasibility problem of finding a point in the intersection of closed constraint sets. It is closely related with Douglas–Rachford method (DR). We prove local convergence of CRM in the same prototypical settings of most theoretical analysis of regular nonconvex DR, whose consideration is made natural by the geometry of the phase retrieval problem. For the purpose, we show that CRM is related to the method of subgradient projections. For many cases when DR is known to converge to a feasible point, we establish that CRM locally provides a better convergence rate. As a root finder, we show that CRM has local convergence whenever Newton–Raphson method does, has quadratic rate whenever Newton–Raphson method does, and exhibits superlinear convergence in many cases when Newton–Raphson method fails to converge at all. We also obtain explicit regions of convergence. As an interesting aside, we demonstrate local convergence of CRM to feasible points in cases when DR converges to fixed points that are not feasible. We demonstrate an extension in higher dimensions, and use it to obtain convergence rate guarantees for sphere and subspace feasibility problems. Armed with these guarantees, we experimentally discover that CRM is highly sensitive to compounding numerical error that may cause it to achieve worse rates than those guaranteed by theory. We then introduce a numerical modification that enables CRM to achieve the theoretically guaranteed rates. Any future works that study CRM for product space formulations of feasibility problems should take note of this sensitivity and account for it in numerical implementations.

On the finite convergence of the Douglas-Rachford algorithm for solving (not necessarily convex) feasibility problems in Euclidean spaces

On the local convergence of the Douglas-Rachford algorithm

Adaptive Douglas–Rachford Algorithms for Biconvex Optimization Problem in the Finite Dimensional Real Hilbert Spaces

Recent Results on Douglas-Rachford Methods for Combinatorial Optimization Problems

Linear Convergence of Subgradient Algorithm for Convex Feasibility on Riemannian Manifolds

Local Convergence Properties of Douglas–Rachford and Alternating Direction Method of Multipliers

The Fun is Finite: Douglas-Rachford and Sudoku Puzzle – Finite Termination and Local Linear Convergence

A parameterized Douglas-Rachford Splitting algorithm for nonconvex optimization

Linear Regularity and Linear Convergence of Projection-Based Methods for Solving Convex Feasibility Problems

Randomized Douglas–Rachford Methods for Linear Systems: Improved Accuracy and Efficiency

PROJECTION ALGORITHMS FOR CONVEX FEASIBILITY PROBLEMS ON HADAMARD MANIFOLDS

New Douglas-Rachford algorithmic structures and their convergence analyses

Douglas-Rachford Feasibility Methods for Matrix Completion Problems

Randomized Douglas-Rachford methods for linear systems: Improved accuracy and efficiency

Convergence depth control for interior point methods

Convergence analysis of the Gauss–Newton method for convex inclusion and convex-composite optimization problems

Understanding the Douglas-Rachford splitting method through the lenses of Moreau-type envelopes

Circumcentering Reflection Methods for Nonconvex Feasibility Problems

On Convergence of the Gauss-Newton Method for Convex Composite Optimization

Convergence Rates of Forward–Douglas–Rachford Splitting Method

Convergence analysis of the DFP algorithm for unconstrained optimization problems on Riemannian manifolds