Federated Learning with Energy-Based Structured Probabilistic Inference
For federated learning practitioners, this addresses the suboptimal aggregation problem in heterogeneous settings, but the improvement is incremental.
The authors propose a CRF-based framework to refine client aggregation weights in federated learning, improving convergence under non-IID heterogeneity. Experiments show consistent performance gains over established baselines.
Federated learning typically aggregates client updates using fixed or heuristic weighting rules, which can be suboptimal when clients have heterogeneous data and varying contributions to the global model. We propose a framework that refines client aggregation weights using Conditional Random Fields (CRFs). Our method defines unary potentials for individual clients and pairwise potentials for all client pairs, allowing the server to model both client-specific reliability and interactions between clients. The resulting CRF inference produces aggregation weights that enable better convergence of the global training objective. Experiments show that, under non-IID heterogeneity, our approach consistently improves performance over well-established federated learning baselines.