Image Classification on CIFAR10 10 clients (test)
92.47AccuracypFedHN-PC
Evaluation Results
| Method | Links | |
|---|---|---|
| pFedHN-PCNumber of clients=10, Backbone=LeNet, Communication steps=5000, Local steps (K)=502021.03 | 92.47 | |
| pFedHNNumber of clients=10, Backbone=LeNet, Communication steps=5000, Local steps (K)=50, Embedding dimension=3.52021.03 | 90.83 | |
| LG-FedAvgNumber of clients=10, Backbone=LeNet, Communication steps=60002021.03 | 89.11 | |
| pFedMeNumber of clients=10, Backbone=LeNet, Communication steps=50002021.03 | 87.69 | |
| FedPerNumber of clients=10, Backbone=LeNet, Communication steps=50002021.03 | 87.27 | |
| LocalNumber of clients=10, Backbone=LeNet, Training steps=2000 optimization steps per client2021.03 | 86.46 | |
| Per-FedAvgNumber of clients=10, Backbone=LeNet, Communication steps=50002021.03 | 76.65 | |
| FedAvgNumber of clients=10, Backbone=LeNet, Communication steps=50002021.03 | 51.42 |