Image Classification on Fashion-MNIST heterogeneous (test)
83Test AccuracyFedLap-Cov
Evaluation results
| Method | Links | ||
|---|---|---|---|
| FedLap-CovCommunication Rounds=50, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 83 | 0.49 | |
| IVON-ADMMCommunication Rounds=50, Posterior Averaging=Full posterior, Backbone=MLP2025.06 | 83 | 0.48 | |
| IVON-ADMM@mgCommunication Rounds=50, Posterior Averaging=Mean weight only, Backbone=MLP2025.06 | 82.9 | 0.48 | |
| FedLapCommunication Rounds=50, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 82.4 | 0.53 | |
| IVON-ADMMCommunication Rounds=25, Posterior Averaging=Full posterior, Backbone=MLP2025.06 | 82.3 | 0.5 | |
| FedDynCommunication Rounds=50, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 82.2 | 0.51 | |
| IVON-ADMM@mgCommunication Rounds=25, Posterior Averaging=Mean weight only, Backbone=MLP2025.06 | 82.2 | 0.5 | |
| FedAvg/FedProxCommunication Rounds=50, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 81.8 | 0.53 | |
| FedDynCommunication Rounds=25, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 81.4 | 0.53 | |
| FedLap-CovCommunication Rounds=25, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 81.3 | 0.54 | |
| FedLapCommunication Rounds=25, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 79.4 | 0.59 | |
| FedAvg/FedProxCommunication Rounds=25, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 78.7 | 0.6 | |
| IVON-ADMM@mgCommunication Rounds=10, Posterior Averaging=Mean weight only, Backbone=MLP2025.06 | 78.7 | 0.62 | |
| IVON-ADMMCommunication Rounds=10, Posterior Averaging=Full posterior, Backbone=MLP2025.06 | 78.7 | 0.62 | |
| FedLap-CovCommunication Rounds=10, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 76.9 | 0.64 | |
| FedDynCommunication Rounds=10, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 75.7 | 0.66 | |
| FedAvg/FedProxCommunication Rounds=10, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 73.9 | 0.73 | |
| FedLapCommunication Rounds=10, Backbone=MLP, Number of Clients=100, Data Distribution=Heterogeneous2025.06 | 73.9 | 0.68 |