Moss: Proxy Model-based Full-Weight Aggregation in Federated Learning with Heterogeneous Models
This addresses inefficiencies in federated learning for heterogeneous mobile devices, offering a novel approach to improve performance, though it appears incremental as it builds on existing FL paradigms.
The paper tackles the problem of sub-optimal accuracy and high training overhead in federated learning with heterogeneous models by proposing Moss, a full-weight aggregation method that accelerates training, reduces on-device time and energy consumption, enhances accuracy, and minimizes network bandwidth usage compared to state-of-the-art baselines.
Modern Federated Learning (FL) has become increasingly essential for handling highly heterogeneous mobile devices. Current approaches adopt a partial model aggregation paradigm that leads to sub-optimal model accuracy and higher training overhead. In this paper, we challenge the prevailing notion of partial-model aggregation and propose a novel "full-weight aggregation" method named Moss, which aggregates all weights within heterogeneous models to preserve comprehensive knowledge. Evaluation across various applications demonstrates that Moss significantly accelerates training, reduces on-device training time and energy consumption, enhances accuracy, and minimizes network bandwidth utilization when compared to state-of-the-art baselines.