Abstract:We propose and analyze a golden ratio primal-dual algorithm for solving structured optimization problems involving the sum of three convex terms - a smooth function with Lipschitzian gradient and two nonsmooth proximal-friendly functions, one of which is composed with a linear mapping. The proposed algorithm is of primal-dual and full-splitting type as it solves the primal and the dual problems simultaneously and does not rely on solving any subproblems or linear system of equations iteratively, the smooth function is handled by gradient evaluation, and the nonsmooth functions are handled by their proximity operators. Several well-known algorithms are closely related, e.g., the classical Arrow-Hurwicz method and the primal-dual algorithm of Chambolle and Pock. In particular, it extends the golden ratio primal-dual algorithm recently proposed by Chang and Yang by including an extra smooth term with Lipschitzian gradient. The convergence rates O(1/N) and O(1/N-2) are established for convex and strongly convex cases, respectively, which differentiate themselves from existing results in terms of the adopted optimality measures. Specifically, to measure optimality, most existing results adopt the primal-dual gap function, a major flaw of which is that it could vanish at nonstationary points. In comparison, we adopt function value residual and feasibility violation as optimality measures, which are conventional for 'constrained optimization. Finally, preliminary numerical results on image reconstruction and elastic net regularization problems are presented to demonstrate the efficiency of the proposed algorithm.

A Primal-Dual Algorithm Framework for Convex Saddle-Point Optimization

An efficient symmetric primal-dual algorithmic framework for saddle point problems

A Primal-Dual Smoothing Framework for Max-Structured Non-Convex Optimization

An Inexact Primal-Dual Smoothing Framework for Large-Scale Non-Bilinear Saddle Point Problems

A convex combination based primal-dual algorithm with linesearch for general convex-concave saddle point problems

Nonlinear preconditioned primal-dual method for a class of structured minimax problems

Primal-Dual Sequential Subspace Optimization for Saddle-point Problems

Accelerated Primal-dual Scheme for a Class of Stochastic Nonconvex-concave Saddle Point Problems

A Second Order Primal–Dual Dynamical System for a Convex–Concave Bilinear Saddle Point Problem

Accelerated Primal-Dual Proximal Gradient Splitting Methods for Convex-Concave Saddle-Point Problems

A splitting primal-dual proximity algorithm for solving composite optimization problems

A New Prediction–Correction Primal–Dual Hybrid Gradient Algorithm for Solving Convex Minimization Problems with Linear Constraints

A Unified Primal-Dual Algorithm Framework for Inequality Constrained Problems

A Modified Primal-Dual Algorithm for Structured Convex Optimization with a Lipschitzian Term

A new primal-dual algorithm for structured convex optimization involving a lipschitzian term

A Smooth Primal-Dual Optimization Framework for Nonsmooth Composite Convex Minimization

A Stochastic Proximal Point Algorithm for Saddle-Point Problems

A Decentralized Proximal Point-type Method for Non-convex Non-concave Saddle Point Problems

Primal-dual proximal bundle and conditional gradient methods for convex problems

Robust Accelerated Primal-Dual Methods for Computing Saddle Points