Finite-sample guarantees for data-driven operator splitting methods via martingale inequalities
Abstract
Operator splitting methods are a fundamental class of algorithms for solving structured monotone inclusion problems arising in optimization, control, and game theory. We consider the case, common in stochastic regimes, where the forward evaluation of one of the constituent operators is either unavailable in closed form or computationally expensive to evaluate, and is therefore approximated using a finite number of noisy oracle samples. We establish distribution-free, finite-sample certificates for the quality of the output produced by data-driven Davis-Yin splitting algorithms. Unlike previous works, our analysis directly controls the residual error via martingale inequalities instead of relying on algorithmic stability arguments for a tailored surrogate loss, yielding the first a priori certificates whose statistical excess provably vanishes with the sample size. We further show that, under linear convergence of the Davis-Yin splitting algorithm, the dependence of our bounds on the iteration count improves from linear growth to exponential decay. We validate our theoretical results on a stochastic portfolio optimization problem with uncertain asset returns.
Related
- Operator Splitting Methods with Online Scaling
- Hopf Algebra of Graphs for Operator Splitting Methods
- Operator splitting methods for non-autonomous evolution equations comprising three parts, with application to the Bloch-Torrey equation of MRI
- On the Local Linear Convergence of Operator Splitting Methods for Conic Programming
- Predictable subordination, sharp martingale inequalities and applications
BibTeX
@misc{martin2026finitesample,
title = {{Finite-sample guarantees for data-driven operator splitting methods via martingale inequalities}},
author = {Andrea Martin and Filippo Fabiani and Giuseppe Belgioioso},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.16353},
eprint = {2609.16353},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.16353},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Finite-sample guarantees for data-driven operator splitting methods via martingale inequalities
AU - Andrea Martin
AU - Filippo Fabiani
AU - Giuseppe Belgioioso
PY - 2026
DA - 2026/09/14
PB - arXiv
DO - 10.48550/arxiv.2609.16353
UR - https://arxiv.org/abs/2609.16353
AB - Operator splitting methods are a fundamental class of algorithms for solving structured monotone inclusion problems arising in optimization, control, and game theory. We consider the case, common in stochastic regimes, where the forward evaluation of one of the constituent operators is either unavailable in closed form or computationally expensive to evaluate, and is therefore approximated using a finite number of noisy oracle samples. We establish distribution-free, finite-sample certificates for the quality of the output produced by data-driven Davis-Yin splitting algorithms. Unlike previous works, our analysis directly controls the residual error via martingale inequalities instead of relying on algorithmic stability arguments for a tailored surrogate loss, yielding the first a priori certificates whose statistical excess provably vanishes with the sample size. We further show that, under linear convergence of the Davis-Yin splitting algorithm, the dependence of our bounds on the iteration count improves from linear growth to exponential decay. We validate our theoretical results on a stochastic portfolio optimization problem with uncertain asset returns.
ER -