Abstract:Federated learning is a distributed paradigm that allows multiple parties to collaboratively train deep learning models without direct exchange of raw data. Nevertheless, the inherent non-independent and identically distributed (non-i.i.d.) nature of data distribution among clients results in significant degradation of the acquired model. The primary goal of this study is to develop a robust federated learning algorithm to address feature shift in clients' samples, potentially arising from a range of factors such as acquisition discrepancies in medical imaging. To reach this goal, we first propose federated feature augmentation ( FedFA $^{l}$ ), a novel feature augmentation technique tailored for federated learning. FedFA $^{l}$ is based on a crucial insight that each client's data distribution can be characterized by first-/second-order statistics ( a.k.a. , mean and standard deviation) of latent features; and it is feasible to manipulate these local statistics globally , i.e. , based on information in the entire federation, to let clients have a better sense of the global distribution across clients. Grounded on this insight, we propose to augment each local feature statistic based on a normal distribution, wherein the mean corresponds to the original statistic, and the variance defines the augmentation scope. Central to FedFA $^{l}$ is the determination of a meaningful Gaussian variance, which is accomplished by taking into account not only biased data of each individual client, but also underlying feature statistics represented by all participating clients. Beyond consideration of low-order statistics in FedFA $^{l}$ , we propose a federated feature alignment component ( FedFA $^{h}$ ) that exploits higher-order feature statistics to gain a more detailed understanding of local feature distribution and enables explicit alignment of augmented features in different clients to promote more consistent feature learning. Combining FedFA $^{l}$ and FedFA $^{h}$ yields our full approach FedFA$+$ . FedFA$+$ is non-parametric, incurs negligible additional communication costs, and can be seamlessly incorporated into popular CNN and Transformer architectures. We offer rigorous theoretical analysis as well as extensive empirical justifications to demonstrate the effectiveness of the algorithm. Our implementation is publicly available at https://github.com/tfzhou/FedFA .

Research on data fusion method based on Federated learning

Research on Multi-source Data Security Fusion Analysis Technology Based on Federated Learning

Federated Learning with Data-Agnostic Distribution Fusion

Analysis of Heterogeneous Data Model Based on Federated Learning

Federated Learning Algorithm Based on Adaptive Gradient Fusion

A Multi-Source Credit Data Fusion Approach Based on Federated Distillation Learning

Research on Federated Learning Algorithms in Non-Independent Identically Distributed Scenarios

Multi-source Data Fusion Base on Block-Term Decomposition in Federated Learning

Towards Faster And Better Federated Learning: A Feature Fusion Approach

Multimodal Fusion with Block Term Decomposition for Asynchronous Federated Learning

Handling Data Heterogeneity in Federated Learning via Knowledge Distillation and Fusion

Federated Learning Model Training Method Based on Data Features Perception Aggregation

FedUC: A Unified Clustering Approach for Hierarchical Federated Learning

A Survey on Federated Learning Technology.

FedCross: Towards Accurate Federated Learning via Multi-Model Cross-Aggregation

Research on Federated Learning Data Management Method Based on Data Lake Technology

Emerging trends in federated learning: from model fusion to federated X learning

Data Augmentation Based Federated Learning

Heterogeneous Federated Learning with Cross-layer Model Fusion.

Federated Feature Augmentation and Alignment

Ensemble Distillation for Robust Model Fusion in Federated Learning