Linearized ADMM for Non-convex Non-smooth Optimization with Convergence Analysis

TL;DR

This paper introduces new linearized ADMM algorithms capable of converging on a broader class of nonconvex, nonsmooth problems, with the ability to update variables in parallel, expanding the applicability of ADMM methods.

Contribution

It proposes two novel linearized ADMM algorithms with convergence guarantees for nonconvex, nonsmooth problems under less restrictive assumptions and supports parallel variable updates.

Findings

Algorithms converge for a broader class of functions

Supports parallel updates of coupled variables

Works for less restrictive nonconvex problems

Abstract

Linearized alternating direction method of multipliers (ADMM) as an extension of ADMM has been widely used to solve linearly constrained problems in signal processing, machine leaning, communications, and many other fields. Despite its broad applications in nonconvex optimization, for a great number of nonconvex and nonsmooth objective functions, its theoretical convergence guarantee is still an open problem. In this paper, we propose a two-block linearized ADMM and a multi-block parallel linearized ADMM for problems with nonconvex and nonsmooth objectives. Mathematically, we present that the algorithms can converge for a broader class of objective functions under less strict assumptions compared with previous works. Furthermore, our proposed algorithm can update coupled variables in parallel and work for less restrictive nonconvex problems, where the traditional ADMM may have difficulties in solving subproblems.