English

A Note on Complexity for Two Classes of Structured Non-Smooth Non-Convex Compositional Optimization

Optimization and Control 2024-11-22 v1

Abstract

This note studies numerical methods for solving compositional optimization problems, where the inner function is smooth, and the outer function is Lipschitz continuous, non-smooth, and non-convex but exhibits one of two special structures that enable the design of efficient first-order methods. In the first structure, the outer function allows for an easily solvable proximal mapping. We demonstrate that, in this case, a smoothing compositional gradient method can find a (δ,ϵ)(\delta,\epsilon)-stationary point--specifically defined for compositional optimization--in O(1/(δϵ2))O(1/(\delta \epsilon^2)) iterations. In the second structure, the outer function is expressed as a difference-of-convex function, where each convex component is simple enough to allow an efficiently solvable proximal linear subproblem. In this case, we show that a prox-linear method can find a nearly ϵ{\epsilon}-critical point in O(1/ϵ2)O(1/\epsilon^2) iterations.

Keywords

Cite

@article{arxiv.2411.14342,
  title  = {A Note on Complexity for Two Classes of Structured Non-Smooth Non-Convex Compositional Optimization},
  author = {Yao Yao and Qihang Lin and Tianbao Yang},
  journal= {arXiv preprint arXiv:2411.14342},
  year   = {2024}
}
R2 v1 2026-06-28T20:08:06.175Z