Complexity of first order inexact Lagrangian and penalty methods for conic convex programming

TL;DR

This paper provides a comprehensive iteration complexity analysis of inexact first order Lagrangian and penalty methods for cone constrained convex problems, detailing their efficiency and computational bounds.

Contribution

It offers new complexity bounds for inexact primal-dual and penalty methods, including cases with and without optimal Lagrange multipliers, using smoothing techniques.

Findings

Complexity of O(1/ε) for inexact augmented Lagrangian methods.

Complexity of O(1/ε^{3/2}) for Nesterov smoothing and penalty methods.

Analysis applicable even when optimal Lagrange multipliers do not exist.

Abstract

In this paper we present a complete iteration complexity analysis of inexact first order Lagrangian and penalty methods for solving cone constrained convex problems that have or may not have optimal Lagrange multipliers that close the duality gap. We first assume the existence of optimal Lagrange multipliers and study primal-dual first order methods based on inexact information and augmented Lagrangian smoothing or Nesterov type smoothing. For inexact (fast) gradient augmented Lagrangian methods we derive a total computational complexity of $\mathcal{O}\left(\frac{1}{\epsilon}\right)$ projections onto a simple primal set in order to attain an $\epsilon-$optimal solution of the conic convex problem. For the inexact fast gradient method combined with Nesterov type smoothing we derive computational complexity $\mathcal{O}\left(\frac{1}{\epsilon^{3/2}}\right)$ projections onto the same set. Then, we assume that optimal Lagrange multipliers for the cone constrained convex problem might not exist, and analyze the fast gradient method for solving penalty reformulations of the problem. For the fast gradient method combined with penalty framework we also derive a total computational complexity of $\mathcal{O}\left(\frac{1}{\epsilon^{3/2}}\right)$ projections onto a simple primal set to attain an $\epsilon-$optimal solution for the original problem.