Sparse Uncertainty-Informed Sampling from Federated Streaming Data

作     者：Röder, Manuel Schleif, Frank-Michael 

作者机构：Technical UAS Würzburg-Schweinfurt Fac. of Computer Science Würzburg Germany Bielefeld University Faculty of Technology Bielefeld Germany Center for Artificial Intelligence and Robotics Würzburg Germany 

出 版 物：《arXiv》 (arXiv)

年 卷 期：2024年

核心收录：

主　　题：Data streams 

摘      要：We present a numerically robust, computationally efficient approach for non-I.I.D. data stream sampling in federated client systems, where resources are limited and labeled data for local model adaptation is sparse and expensive. The proposed method identifies relevant stream observations to optimize the underlying client model, given a local labeling budget, and performs instantaneous labeling decisions without relying on any memory buffering strategies. Our experiments show enhanced training batch diversity and an improved numerical robustness of the proposal compared to existing strategies over large-scale data streams, making our approach an effective and convenient solution in FL environments. Copyright © 2024, The Authors. All rights reserved.