Convergence Rates for Stochastic Proximal and Projection Estimators
Diego Morales, Pedro Pérez-Aros, Emilio Vilches · 2026
Read Convergence Rates for Stochastic Proximal and Projection Estimators by Diego Morales, Pedro Pérez-Aros, Emilio Vilches (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-06750.txt.
Description
Diego Morales, Pedro Pérez-Aros, and Emilio Vilches establish how fast certain smoothed estimators close in on the objects they approximate. The estimators come from smooth approximations of infimal convolutions, a way of blending two functions. They are used to stand in for proximal mappings and metric projections, two workhorse operations in optimization. The paper's goal is explicit rates with explicit constants.
For weakly convex, possibly nonsmooth problems, the authors prove a square-root rate that spells out the dependence on dimension. Under stronger smoothness assumptions the rate improves to linear, again with constants in plain view, and projection estimates are refined for sets with smooth boundaries. Examples included in the paper show these rates cannot be improved in general.
Catalog record
| CREATOR | Diego Morales, Pedro Pérez-Aros, Emilio Vilches |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602-06750