aiqbench analysis
This paper addresses the convergence and feasibility challenges in functionally constrained federated learning by introducing a variance-reducing framework that eliminates dual-variable tuning. The authors demonstrate how their primal-only updates and error feedback mechanisms maintain stationarity guarantees under stochastic gradients and partial client participation. These theoretical advances are validated through experiments on fairness-constrained classification and deep reinforcement learning tasks.
Variance Reduction for Stochastic Switching in Functionally Constrained Federated Optimization
arXiv:2601.16897v2 Announce Type: replace Abstract: Federated learning (FL) is increasingly used in safety-critical and fairness-sensitive applications, where learning objectives must be optimized subject to functional constraints. This setting is challenging due to costly projection onto the resulting feasible set, stochastic constraint information, partial client participation, and limited communication at edge devices. We propose a federated stochastic switching framework based on primal-only updates, avoiding dual-variable tuning for functionally constrained federated optimization. For convex objectives, the framework further supports compression with error feedback to control compression-induced bias. We establish high-probability convergence and feasibility guarantees under stochastic gradients and partial client participation, and extend the analysis to weakly convex objectives through proximal-stationarity guarantees. Our bounds isolate the effects of stochastic gradients, constraint-evaluation noise, and partial participation, and show that direct switching can retain a nonvanishing constraint-estimation error even with exact local constraint evaluations. To overcome this limitation, we develop variance-reducing constraint estimators that yield vanishing feasibility and stationarity guarantees with fixed per-round client participation and constraint-evaluation batch sizes. Experiments on Neyman-Pearson classification, fairness-constrained learning on the ADULT dataset, and constrained deep reinforcement learning demonstrate the effectiveness of the proposed framework.
Abstract:Federated learning (FL) is increasingly used in safety-critical and fairness-sensitive applications, where learning objectives must be optimized subject to functional constraints. This setting is challenging due to costly projection onto the resulting feasible set, stochastic constraint information, partial client participation, and limited communication at edge devices. We propose a federated stochastic switching framework based on primal-only updates, avoiding dual-variable tuning for functionally constrained federated optimization. For convex objectives, the framework further supports compression with error feedback to control compression-induced bias. We establish high-probability convergence and feasibility guarantees under stochastic gradients and partial client participation, and extend the analysis to weakly convex objectives through proximal-stationarity guarantees. Our bounds isolate the effects of stochastic gradients, constraint-evaluation noise, and partial participation, and show that direct switching can retain a nonvanishing constraint-estimation error even with exact local constraint evaluations. To overcome this limitation, we develop variance-reducing constraint estimators that yield vanishing feasibility and stationarity guarantees with fixed per-round client participation and constraint-evaluation batch sizes. Experiments on Neyman-Pearson classification, fairness-constrained learning on the ADULT dataset, and constrained deep reinforcement learning demonstrate the effectiveness of the proposed framework.
Machine Learning (cs.LG); Optimization and Control (math.OC); Machine Learning (stat.ML)
arXiv:2601.16897 [cs.LG]
(or arXiv:2601.16897v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2601.16897 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 281 words
Related coverage
GEAK V3: Agent-Driven, Repository-Level GPU Kernel Optimization across HIP, Triton, and FlyDSL on AMD GPUs
AMD · Jul 20, 2026
Bayesian Optimization on Function Spaces via Sparse RKHS Manifolds
arXiv · Oct 7, 2026
Decentralized Projection-free Online Upper-Linearizable Optimization with Applications to DR-Submodular Optimization
arXiv · Oct 1, 2026
Researchers Push Quantum Conference Key Agreement (QCKA) Past 20 per cent.
Quantum Zeitgeist · Jul 19, 2026