Advances in Federated Learning and Dynamic Systems

The field of federated learning and dynamic systems is rapidly evolving, with a focus on improving performance, fairness, and robustness in non-IID settings. Researchers are exploring innovative approaches to address the challenges of heterogeneous data distributions, quantity skew, and time delays. Notable developments include the integration of adaptive boosting mechanisms, clustered federated learning, and personalized prototype learning. These advancements have the potential to significantly improve the accuracy and reliability of models in real-world applications.

Some noteworthy papers in this area include: FeDABoost, which proposes a novel FL framework that integrates a dynamic boosting mechanism and an adaptive gradient aggregation strategy. CORNFLQS, which presents a robust clustered federated learning approach for non-IID data with quantity skew. DPMM-CFL, which introduces a nonparametric Bayesian inference approach to jointly infer the number of clusters and client assignments in clustered federated learning.

Sources

FeDABoost: Fairness Aware Federated Learning with Adaptive Boosting

Eigenvalue Tracking of Large-Scale Systems Impacted by Time Delays

A Robust Clustered Federated Learning Approach for Non-IID Data with Quantity Skew

Personalized federated prototype learning in mixed heterogeneous data scenarios

Technical note on Fisher Information for Robust Federated Cross-Validation

On the Duality Between Quantized Time and States in Dynamic Simulation

On Provable Benefits of Muon in Federated Learning

Adaptive Federated Learning via Dynamical System Model

Machine Unlearning in Speech Emotion Recognition via Forget Set Alone

Federated Self-Supervised Learning for Automatic Modulation Classification under Non-IID and Class-Imbalanced Data

Power Transform Revisited: Numerically Stable, and Federated

Federated Unlearning in the Wild: Rethinking Fairness and Data Discrepancy

Validation of Various Normalization Methods for Brain Tumor Segmentation: Can Federated Learning Overcome This Heterogeneity?

DPMM-CFL: Clustered Federated Learning via Dirichlet Process Mixture Model Nonparametric Clustering