FedPURIN: Programmed Update and Reduced INformation for Sparse Personalized Federated Learning

2025-10-16

AI Summary

To address the high communication overhead and deployment constraints in personalized federated learning (PFL) caused by statistical heterogeneity, this paper proposes FedPURIN—a novel framework that introduces integer programming into the PFL parameter update mechanism for the first time. By modeling and optimizing parameter decoupling, FedPURIN precisely identifies critical parameter subsets, enabling programmable sparse updates and aggregation. Its core innovation lies in establishing a sparsification paradigm that jointly optimizes communication efficiency and personalized model performance. Evaluated on standard image classification benchmarks under non-IID data settings, FedPURIN achieves state-of-the-art (SOTA) model accuracy while significantly reducing communication volume. Experimental results demonstrate its practicality and scalability in edge intelligence scenarios, where bandwidth and device resource constraints are stringent.

Technology Category

• Machine Learning: Distributed Machine Learning & Federated Learning
• Natural Language Processing: Learning & Optimization for NLP
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Personalized Federated Learning (PFL) has emerged as a critical research frontier addressing data heterogeneity issue across distributed clients. Novel model architectures and collaboration mechanisms are engineered to accommodate statistical disparities while producing client-specific models. Parameter decoupling represents a promising paradigm for maintaining model performance in PFL frameworks. However, the communication efficiency of many existing methods remains suboptimal, sustaining substantial communication burdens that impede practical deployment. To bridge this gap, we propose Federated Learning with Programmed Update and Reduced INformation (FedPURIN), a novel framework that strategically identifies critical parameters for transmission through an integer programming formulation. This mathematically grounded strategy is seamlessly integrated into a sparse aggregation scheme, achieving a significant communication reduction while preserving the efficacy. Comprehensive evaluations on standard image classification benchmarks under varied non-IID conditions demonstrate competitive performance relative to state-of-the-art methods, coupled with quantifiable communication reduction through sparse aggregation. The framework establishes a new paradigm for communication-efficient PFL, particularly advantageous for edge intelligence systems operating with heterogeneous data sources.

Problem

• Addressing data heterogeneity in personalized federated learning
• Reducing communication burdens through sparse parameter transmission
• Improving efficiency for edge intelligence with heterogeneous data

Innovation

• Programmed parameter selection via integer programming
• Sparse aggregation scheme reduces communication burden
• Maintains model efficacy with quantifiable communication reduction