Privacy-Preserving Aggregation for Decentralized Learning with Byzantine-Robustness

arXiv:2404.17970v1 Announce Type: new
Abstract: Decentralized machine learning (DL) has been receiving an increasing interest recently due to the elimination of a single point of failure, present in Federated learning setting. Yet, it is threatened by the looming threat of Byzantine clients who intentionally disrupt the learning process by broadcasting arbitrary model updates to other clients, seeking to degrade the performance of the global model. In response, robust aggregation schemes have emerged as promising solutions to defend against such Byzantine …

aggregation arxiv clients cs.ai cs.cr decentralized disrupt failure federated federated learning interest machine machine learning point privacy process robustness single threat updates