FedPrism: Adaptive Personalized Federated Learning under Non-IID Data

2026-03-09

AI Summary

This work addresses the challenge of balancing global generalization and local personalization in federated learning under non-IID data, where statistical heterogeneity often degrades performance. To this end, we propose FedPrism, a novel framework that introduces a Prism decomposition mechanism to decouple each client’s model into three components: a global base, a shared group-specific module, and a private local part. FedPrism integrates adaptive client clustering with a dual-stream architecture that dynamically routes prediction paths based on the confidence of local experts. This approach effectively reconciles the tension between generalization and personalization, significantly outperforming existing static aggregation and hard-clustering baselines in highly heterogeneous non-IID settings, while demonstrating superior accuracy, robustness, and adaptability.

Technology Category

Machine Learning: Distributed Machine Learning & Federated LearningSearch and Optimization: Local SearchMultiagent Systems: Adversarial Agents

Application Category

User Modeling, Personalization and Recommendation: Federated recommendation systems and personalizationSystems and Infrastructure for Web, Mobile and WoT: Decentralized Web and Fediverse systemsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Federated Learning (FL) suffers significant performance degradation in real-world deployments characterized by moderate to extreme statistical heterogeneity (non-IID client data). While global aggregation strategies promote broad generalization, they often fail to capture the diversity of local data distributions, leading to suboptimal personalization. We address this problem with FedPrism, a framework that uses two main strategies. First, it uses a Prism Decomposition method that builds each client's model from three parts: a global foundation, a shared group part for similar clients, and a private part for unique local data. This allows the system to group similar users together automatically and adapt if their data changes. Second, we include a Dual-Stream design that runs a general model alongside a local specialist. The system routes predictions between the general model and the local specialist based on the specialist's confidence. Through systematic experiments on non-IID data partitions, we demonstrate that FedPrism exceeds static aggregation and hard-clustering baselines, achieving significant accuracy gains under high heterogeneity. These results establish FedPrism as a robust and flexible solution for federated learning in heterogeneous environments, effectively balancing generalizable knowledge with adaptive personalization.

Problem

• Federated Learning
• Non-IID Data
• Personalization
• Statistical Heterogeneity
• Model Generalization

Innovation

• Federated Learning
• Personalization
• Non-IID Data
• Model Decomposition
• Dual-Stream Architecture