Abstract:In this paper, we establish the convergence of the proximal alternating direction method of multipliers (ADMM) and block coordinate descent (BCD) method for nonseparable minimization models with quadratic coupling terms. The novel convergence results presented in this paper answer several open questions that have been the subject of considerable discussion. We firstly extend the 2-block proximal ADMM to linearly constrained convex optimization with a coupled quadratic objective function, an area where theoretical understanding is currently lacking, and prove that the sequence generated by the proximal ADMM converges in point-wise manner to a primal-dual solution pair. Moreover, we apply randomly permuted ADMM (RPADMM) to nonseparable multi-block convex optimization, and prove its expected convergence for a class of nonseparable quadratic programming problems. When the linear constraint vanishes, the 2-block proximal ADMM and RPADMM reduce to the 2-block cyclic proximal BCD method and randomly permuted BCD (RPBCD). Our study provides the first iterate convergence result for 2-block cyclic proximal BCD without assuming the boundedness of the iterates. We also theoretically establish the expected iterate convergence result concerning multi-block RPBCD for convex quadratic optimization. In addition, we demonstrate that RPBCD may have a worse convergence rate than cyclic proximal BCD for 2-block convex quadratic minimization problems. Although the results on RPADMM and RPBCD are restricted to quadratic minimization models, they provide some interesting insights: (1) random permutation makes ADMM and BCD more robust for multi-block convex minimization problems; (2) cyclic BCD may outperform RPBCD for “nice” problems, and RPBCD should be applied with caution when solving general convex optimization problems especially with a few blocks.

A Note on the Sufficient Initial Condition Ensuring the Convergence of Directly Extended 3-Block ADMM for Special Semidefinite Programming

An Extended ADMM for 3-Block Nonconvex Nonseparable Problems with Applications

The symmetric ADMM with indefinite proximal regularization and its application

A Convergent $3$-Block Semi-Proximal ADMM for Convex Minimization Problems with One Strongly Convex Block

Sequential inertial linear ADMM algorithm for nonconvex and nonsmooth multiblock problems with nonseparable structure

The Direct Extension of ADMM for Multi-Block Convex Minimization Problems is Not Necessarily Convergent

An Efficient Inexact Symmetric Gauss-Seidel Based Majorized ADMM for High-Dimensional Convex Composite Conic Programming

Linearized ADMM for Nonconvex Nonsmooth Optimization with Convergence Analysis

On the Convergence Analysis of the Alternating Direction Method of Multipliers with Three Blocks

Convergent Prediction-Correction-based ADMM for Multi-Block Separable Convex Programming

Efficient Semidefinite Programming with approximate ADMM

A Bregman-style Partially Symmetric Alternating Direction Method of Multipliers for Nonconvex Multi-block Optimization

Extended ADMM and BCD for Nonseparable Convex Minimization Models with Quadratic Coupling Terms: Convergence Analysis and Insights

A Unified Alternating Direction Method of Multipliers by Majorization Minimization

A Bregman-Style Improved ADMM and its Linearized Version in the Nonconvex Setting: Convergence and Rate Analyses

Parallel Direction Method of Multipliers

Convergence Analysis of Generalized ADMM with Majorization for Linearly Constrained Composite Convex Optimization

Two-block vs. Multi-block ADMM: An empirical evaluation of convergence

Stochastic linearized generalized alternating direction method of multipliers: Expected convergence rates and large deviation properties

A Generalization of Linearized Alternating Direction Method of Multipliers for Solving Two-Block Separable Convex Programming

Linear Convergence of the Alternating Direction Method of Multipliers for a Class of Convex Optimization Problems