| q-FFL |
Paper |
Code |
CMU CS |
q-Fair Federated Learning (q-FFL), a novel optimization objective inspired by fair resource allocation in wireless networks that encourages a more fair (specifically, a more uniform) accuracy distribution across devices in federated networks |
ICLR 2020 |
| FedProf |
Paper |
Code |
Weiwei Lin |
Based on a dynamic data representation profiling and matching scheme, a selective FL training algorithm FedProf is proposed which adaptively adjusts clients’ participation chance based on their profile dissimilarity. |
Not published yet |
| Optimal Client Sampling |
Paper |
Code |
Cambridge & KAUST |
A principled optimal client sampling scheme, capable of identifying the most informative clients in any given communication round. It works by minimizing the variance of the stochastic gradient produced by the partial participation procedure, which then translates to a reduction in the number of communication rounds. |
NIPS2020 Workshop |
| Power-Of-Choice |
Paper |
No code released |
CMU ECE |
A communication- and computation-efficient client selection framework that can flexibly span the trade-off between convergence speed and solution bias. (different from unbiased client participation, where clients are selected at random or in proportion of their data sizes.) |
Not published yet |
| FedCS |
Paper |
No code released |
Kyoto University |
The earliest client selection study, in which Nishio et al. proposed a rather intuitive selection scheme and highlighted the importance of client update number to the model performance. Assuming that the training status of clients (e.g. the training time, resource usage, etc) are known or can at least be calculated. |
ICC 2019 |
| MAB-based Client Selection |
Paper Video |
No code released |
Kyoto University |
When the computation and communication resource of clients cannot be estimated: A multi-armed bandit (MAB)-based client selection method to solve the exploration and exploitation trade-off and reduce the time consumption for FL in mobile networks. |
IEEE GLOBECOM 2020 |
| CS-UCB-Q |
Paper |
Code |
Singapore University of Technology and Design |
A multi-armed bandit-based framework for online client scheduling (CS) in FL without knowing wireless channel state information and statistical characteristics of clients. |
Not published yet |
| Oort |
Paper |
Code |
University of Michigan |
Oort is proposed to improve the performance of federated training and testing with guided participant selection. With an aim to improve time-to-accuracy performance in model training, Oort prioritizes the use of those clients who have both data that offers the greatest utility in improving model accuracy and the capability to run training quickly. To enable FL developers to interpret their results in model testing, Oort enforces their requirements on the distribution of participant data while improving the duration of federated testing by cherry-picking clients. |
Distinguished Artifact Award at OSDI'2021 |