Image Classification on CIFAR-10 Practical heterogeneous (alpha = 0.1, K = 20) (test)
91.15AccuracyCAFedCL
Evaluation results
| Method | Links | ||
|---|---|---|---|
| CAFedCLBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 91.15 | 4.87 | |
| FedTGPBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 89.24 | 7.03 | |
| FedProtoBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 88.33 | 8.13 | |
| FedRCLBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 87.26 | 8.79 | |
| MP-FedCLBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 86.94 | 8.46 | |
| MOONBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 82.06 | 11.19 | |
| FedLCBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 78.48 | 7.1 | |
| FedProxBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 61.82 | 10.07 | |
| FedAvgBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 60.32 | 12.03 | |
| FedProcBackbone=ResNet-18, Normalization=Group Normalization, Local Training Epochs=5, Optimizer=SGD, Number of runs=52026.03 | 44.53 | 6.17 |