Optimization Online


Hybrid Middle Proximal ADMM for Linearly Constrained Convex Optimization

Jianchao Bai (bjc1987***at***163.com)

Abstract: This paper focuses on a family of Middle Proximal Alternating Direction Method of Multipliers (MP-ADMM) for solving multi-block separable convex optimization subject to linearly constraints and structured constraints. This one-parameter family of MP-ADMM combines both Jacobi and Gauss-Seidel type of ADMM and proximal point techniques are only applied to the middle subproblems to promote its convergence. We analyze the global convergence as well as the worst-case ${\mathcal O} (1/t)$ ergodic convergence rate of this new algorithm, where $t$ is the iteration number. Moreover, linear convergence of the algorithm is also established under the assumption that the sub-differential of each component function in the objective function is piece-wise linear. {\color{red}Besides, a linearized version of MP-ADMM is also discussed to accelerate the convergence.} Compared with several state-of-the-art algorithms in the literature, this new MP-ADMM performs very well for solving a latent variable Gaussian graphical model selection problem arising in statistical learning.

Keywords: Alternating direction method of multipliers; Combination technique; Proximal term; Variational inequality; Global convergence

Category 1: Convex and Nonsmooth Optimization (Convex Optimization )

Citation: submitted

Download: [PDF]

Entry Submitted: 11/09/2017
Entry Accepted: 11/09/2017
Entry Last Modified: 02/07/2018

Modify/Update this entry

  Visitors Authors More about us Links
  Subscribe, Unsubscribe
Digest Archive
Search, Browse the Repository


Coordinator's Board
Classification Scheme
Give us feedback
Optimization Journals, Sites, Societies
Mathematical Optimization Society