Distributed Learning with a Local Touch: Improving Efficiency in Multiparty Learning

Exploring the advancements in multiparty learning, this work highlights various frameworks and applications. It reviews a secure federated learning framework, discusses a scalable system for mobile devices, and presents strategies to reduce communication costs. Additionally, it introduces a new protocol, FedCS, for managing computing resources and addresses challenges in learning non-IID data by proposing a globally shared data subset. Lastly, it features a novel optimization method, MOCHA, for federated multi-task learning.