Decentralized machine learning has broadened its scope recently with the invention of Federated Learning (FL), Split Learning (SL), and their hybrids like Split Federated Learning (SplitFed or SFL). The goal of SFL is to reduce the computational power required by each client in FL and parallelize SL while maintaining privacy. This paper investigates the robustness of SFL against packet loss on communication links. The performance of various SFL aggregation strategies is examined by splitting the model at two points -- shallow split and deep split -- and testing whether the split point makes a statistically significant difference to the accuracy of the final model. Experiments are carried out on a segmentation model for human embryo images and indicate the statistically significant advantage of a deeper split point.
@article{arxiv.2307.13851,
title = {SplitFed resilience to packet loss: Where to split, that is the question},
author = {Chamani Shiranthika and Zahra Hafezi Kafshgari and Parvaneh Saeedi and Ivan V. Bajić},
journal= {arXiv preprint arXiv:2307.13851},
year = {2023}
}
Comments
10 pages, 4 figures, MICCAI 2023 Workshop on Distributed, Collaborative and Federated Learning