Image Classification on CIFAR-10 Cross-Device alpha=0.5
66.94Top-1 AccuracyMOON + CONST
Evaluation results
| Method | Links | |
|---|---|---|
| MOON + CONSTModel=ResNet-182026.06 | 66.94 | |
| FEDAVG + CONSTModel=ResNet-182026.06 | 66.51 | |
| FEDDYN + CONSTModel=ResNet-182026.06 | 64.29 | |
| FEDPROX + CONSTModel=ResNet-182026.06 | 63.51 | |
| SCAFFOLD + CONSTModel=ResNet-182026.06 | 63.49 | |
| FEDSAM + CONSTModel=ResNet-182026.06 | 63.45 | |
| FEDSAMModel=ResNet-182026.06 | 62.52 | |
| MOONModel=ResNet-182026.06 | 57.84 | |
| FEDPROXModel=ResNet-182026.06 | 56.79 | |
| SCAFFOLDModel=ResNet-182026.06 | 56.47 | |
| FEDAVG + CONSTModel=LeNet-52026.06 | 54.28 | |
| FEDDYN + CONSTModel=LeNet-52026.06 | 54.07 | |
| FEDAVGModel=ResNet-182026.06 | 54.07 | |
| SCAFFOLD + CONSTModel=LeNet-52026.06 | 53.82 | |
| FEDPROX + CONSTModel=LeNet-52026.06 | 53.09 | |
| FEDDYNModel=ResNet-182026.06 | 52.64 | |
| MOON + CONSTModel=LeNet-52026.06 | 48.66 | |
| FEDAVGModel=LeNet-52026.06 | 46.12 | |
| SCAFFOLDModel=LeNet-52026.06 | 45.66 | |
| FEDPROXModel=LeNet-52026.06 | 45.58 | |
| FEDDYNModel=LeNet-52026.06 | 44.93 | |
| MOONModel=LeNet-52026.06 | 43.89 |