Federated Submodel Optimization for Hot and Cold Data Features
This addresses efficiency issues in federated learning for applications with sparse data, though it is an incremental improvement over existing methods.
The paper tackles the slowdown in federated learning caused by sparse, non-i.i.d. data by proposing Federated Submodel Averaging (FedSubAvg), which ensures accurate global updates and achieves significant performance improvements over FedAvg and its variants on public and industrial datasets.
We study practical data characteristics underlying federated learning, where non-i.i.d. data from clients have sparse features, and a certain client's local data normally involves only a small part of the full model, called a submodel. Due to data sparsity, the classical federated averaging (FedAvg) algorithm or its variants will be severely slowed down, because when updating the global model, each client's zero update of the full model excluding its submodel is inaccurately aggregated. Therefore, we propose federated submodel averaging (FedSubAvg), ensuring that the expectation of the global update of each model parameter is equal to the average of the local updates of the clients who involve it. We theoretically proved the convergence rate of FedSubAvg by deriving an upper bound under a new metric called the element-wise gradient norm. In particular, this new metric can characterize the convergence of federated optimization over sparse data, while the conventional metric of squared gradient norm used in FedAvg and its variants cannot. We extensively evaluated FedSubAvg over both public and industrial datasets. The evaluation results demonstrate that FedSubAvg significantly outperforms FedAvg and its variants.