Image Classification on DermaMNIST Mild Non-IID alpha=1.0
74.32AccuracyCentralized
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| CentralizedTraining Mode=Centralized2026.05 | 74.32 | 49.72 | 4.44 | |
| ChainLearnTraining Mode=Federated Ensemble, Dropout Simulation=False2026.05 | 74.08 | 46.26 | 4.74 | |
| EqualWt-EnsTraining Mode=Federated Ensemble2026.05 | 73.78 | 46.28 | 4.74 | |
| ChainLearn-DropoutTraining Mode=Federated Ensemble, Dropout Simulation=True2026.05 | 73.64 | 45.84 | 3.41 | |
| FedMDTraining Mode=Federated2026.05 | 71.88 | 38.28 | 10.59 | |
| Local-BestTraining Mode=Local2026.05 | 70.91 | 37.24 | 19.53 | |
| FedAvgTraining Mode=Federated2026.05 | 69.95 | 28.94 | 4.6 | |
| FedProxTraining Mode=Federated2026.05 | 69.31 | 26.7 | 5.89 |