FedSI: Federated Subnetwork Inference for Efficient Uncertainty Quantification
This addresses the need for scalable uncertainty quantification in federated learning, particularly for applications with data heterogeneity, though it appears incremental as it builds on existing Bayesian and subnetwork approaches.
The paper tackles the problem of efficient uncertainty quantification in personalized federated learning by proposing FedSI, a Bayesian deep neural network framework that uses client-specific subnetwork inference. It shows improved performance over existing methods on three benchmark datasets in heterogeneous scenarios.
While deep neural networks (DNNs) based personalized federated learning (PFL) is demanding for addressing data heterogeneity and shows promising performance, existing methods for federated learning (FL) suffer from efficient systematic uncertainty quantification. The Bayesian DNNs-based PFL is usually questioned of either over-simplified model structures or high computational and memory costs. In this paper, we introduce FedSI, a novel Bayesian DNNs-based subnetwork inference PFL framework. FedSI is simple and scalable by leveraging Bayesian methods to incorporate systematic uncertainties effectively. It implements a client-specific subnetwork inference mechanism, selects network parameters with large variance to be inferred through posterior distributions, and fixes the rest as deterministic ones. FedSI achieves fast and scalable inference while preserving the systematic uncertainties to the fullest extent. Extensive experiments on three different benchmark datasets demonstrate that FedSI outperforms existing Bayesian and non-Bayesian FL baselines in heterogeneous FL scenarios.