English

Asymptotic linear convergence of fully-corrective generalized conditional gradient methods

Optimization and Control 2023-07-17 v2

Abstract

We propose a fully-corrective generalized conditional gradient method (FC-GCG) for the minimization of the sum of a smooth, convex loss function and a convex one-homogeneous regularizer over a Banach space. The algorithm relies on the mutual update of a finite set Ak\mathcal{A}_k of extremal points of the unit ball of the regularizer and of an iterate ukcone(Ak)u_k \in \operatorname{cone}(\mathcal{A}_k). Each iteration requires the solution of one linear problem to update Ak\mathcal{A}_k and of one finite dimensional convex minimization problem to update the iterate. Under standard hypotheses on the minimization problem we show that the algorithm converges sublinearly to a solution. Subsequently, imposing additional assumptions on the associated dual variables, this is improved to a linear rate of convergence. The proof of both results relies on two key observations: First, we prove the equivalence of the considered problem to the minimization of a lifted functional over a particular space of Radon measures using Choquet's theorem. Second, the FC-GCG algorithm is connected to a Primal-Dual-Active-point Method (PDAP) on the lifted problem for which we finally derive the desired convergence rates.

Keywords

Cite

@article{arxiv.2110.06756,
  title  = {Asymptotic linear convergence of fully-corrective generalized conditional gradient methods},
  author = {Kristian Bredies and Marcello Carioni and Silvio Fanzon and Daniel Walter},
  journal= {arXiv preprint arXiv:2110.06756},
  year   = {2023}
}

Comments

50 pages, 3 figures

R2 v1 2026-06-24T06:51:40.242Z