Cross-Silo Federated Learning for Multi-Tier Networks with Vertical and Horizontal Data Partitioning

arXiv:2108.08930v4 Announce Type: replace
Abstract: We consider federated learning in tiered communication networks. Our network model consists of a set of silos, each holding a vertical partition of the data. Each silo contains a hub and a set of clients, with the silo's vertical data shard partitioned horizontally across its clients. We propose Tiered Decentralized Coordinate Descent (TDCD), a communication-efficient decentralized training algorithm for such two-tiered networks. The clients in each silo perform multiple local gradient steps before sharing updates …

abstract arxiv communication cs.dc cs.lg data data partitioning federated learning hub network networks partitioning set type