English

New operator designs for Halpern iterations with explicit rates under H\"older error bounds

Optimization and Control 2026-05-29 v2

Abstract

We investigate the asymptotic behavior of Halpern-type iterations applied to quasi-nonexpansive operators arising in best approximation problems over the intersection of finitely many closed convex sets in Rn\mathbb{R}^n. Assuming a local decrease condition for the underlying operator and standard requirements on the stepsizes (αk)(0,1)(\alpha_k) \subset (0,1), we first prove strong convergence of the Halpern sequence xk+1=αkx0+(1αk)Txkx_{k+1} = \alpha_k x_0 + (1-\alpha_k) T x_k to the best approximation point xx^\star in the intersection set, that is, the metric projection of x0x_0 onto that set. Under the additional assumption that the intersection satisfies a H\"older-type error bound with exponent γ(0,1]\gamma \in (0,1], we then derive explicit convergence rates for both feasibility and norm error: the distance from xkx_k to the intersection set decays like O(αkγ/(2γ))\mathcal O(\alpha_k^{\gamma/(2-\gamma)}), while the norm error xkx\|x_k - x^\star\| decays like O(αkγ/(42γ))\mathcal O(\alpha_k^{\gamma/(4-2\gamma)}). These results apply to most projection-type operators used in convex feasibility problems (including MAP, CRM/SCCRM, Cimmino and 3PM/A3PM) and extend classical convergence analyses of the Halpern-type iterations by providing explicit, geometry-dependent rates governed by H\"older-type error bounds. Our numerical experiments show that Halpern-type iterations combined with most of these projection-type operators are quicker than Dykstra's algorithm to find the projection of a point in an intersection of ellipsoids or in an intersection of polyhedrons.

Keywords

Cite

@article{arxiv.2601.14451,
  title  = {New operator designs for Halpern iterations with explicit rates under H\"older error bounds},
  author = {Pablo Barros and Vincent Guigues and Roger Behling and Luiz-Rafael Santos},
  journal= {arXiv preprint arXiv:2601.14451},
  year   = {2026}
}
R2 v1 2026-07-01T09:13:12.712Z