Hierarchical Federated ADMM
This work addresses privacy and performance issues in federated learning for distributed systems, representing an incremental improvement by adapting ADMM to hierarchical settings.
The authors tackled the problem of hierarchical federated learning by introducing a novel framework based on the alternating direction method of multipliers (ADMM), which enhances privacy and demonstrates superior learning convergence and accuracy compared to conventional gradient descent-based methods.
In this paper, we depart from the widely-used gradient descent-based hierarchical federated learning (FL) algorithms to develop a novel hierarchical FL framework based on the alternating direction method of multipliers (ADMM). Within this framework, we propose two novel FL algorithms, which both use ADMM in the top layer: one that employs ADMM in the lower layer and another that uses the conventional gradient descent-based approach. The proposed framework enhances privacy, and experiments demonstrate the superiority of the proposed algorithms compared to the conventional algorithms in terms of learning convergence and accuracy. Additionally, gradient descent on the lower layer performs well even if the number of local steps is very limited, while ADMM on both layers lead to better performance otherwise.