Abstract:We consider a class of constrained optimization problems where the objective function is a sum of a smooth function and a nonconvex non-Lipschitz function. Many problems in sparse portfolio selection, edge preserving image restoration, and signal processing can be modelled in this form. First, we propose the concept of the Karush-Kuhn-Tucker (KKT) stationary condition for the non-Lipschitz problem and show that it is necessary for optimality under a constraint qualification called the relaxed constant positive linear dependence (RCPLD) condition, which is weaker than the Mangasarian-Fromovitz constraint qualification and holds automatically if all the constraint functions are affine. Then we propose an augmented Lagrangian (AL) method in which the augmented Lagrangian subproblems are solved by a nonmonotone proximal gradient method. Under the assumption that a feasible point is known, we show that any accumulation point of the sequence generated by our method must be a feasible point. Moreover, if RCPLD holds at such an accumulation point, then it is a KKT point of the original problem. Finally, we conduct numerical experiments to compare the performance of our AL method and the interior point (IP) method for solving two sparse portfolio selection models. The numerical results demonstrate that our method is not only comparable to the IP method in terms of solution quality, but also substantially faster than the IP method.

Augmented Lagrangian method for nonlinear circular conic programs: a local convergence analysis

Convergence analysis of the augmented Lagrangian method for nonlinear second-order cone optimization problems

Convergence of the Augmented Lagrangian Method for Nonlinear Optimization Problems over Second-Order Cones

Convergence Analysis of the Augmented Lagrangian Method for ℓ _p -Norm Cone Optimization Problems with P ≥ 2

Inexact Augmented Lagrangian Methods for Conic Programs: Quadratic Growth and Linear Convergence

The Rate of Convergence of the Augmented Lagrangian Method for Nonlinear Semidefinite Programming.

On the approximate augmented Lagrangian for nonlinear symmetric cone programming

Exploiting cone approximations in an augmented Lagrangian method for conic optimization

A class of nonlinear Lagrangians for nonconvex second order cone programming

The Augmented Lagrangian Method for a Type of Inverse Quadratic Programming Problems over Second-Order Cones

A Unified Convergence Analysis of a Second-Order Method of Multipliers for Nonlinear Conic Programming

A Smoothing Augmented Lagrangian Method for Nonconvex, Nonsmooth Constrained Programs and Its Applications to Bilevel Problems

Global Convergence Analysis of the Power Proximal Point and Augmented Lagrangian Method

Convergence Analysis of a Nonlinear Lagrange Algorithm for Nonconvex Semidefinite Programming

An Augmented Lagrangian Method for State Constrained Linear Parabolic Optimal Control Problems

First-Order Primal-Dual Augmented Lagrangian Method for Nonlinear Cone Constrained Composite Convex Optimization

An Augmented Lagrangian Method for Non-Lipschitz Nonconvex Programming

Log-Sigmoid Nonlinear Lagrange Method for Nonlinear Optimization Problems over Second-Order Cones

The Rate of Convergence of the Augmented Lagrangian Method for a Nonlinear Semidefinite Nuclear Norm Composite Optimization Problem

Local convergence of augmented Lagrangian methods for composite optimization

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