Image Classification on FashionMNIST (IID)
90.6AccuracyFedAvg
Evaluation Results
| Method | Links | |
|---|---|---|
| FedAvgCluster size Nm=10, Local epoch K=5, Mini-batch size=64, Backbone=6-layer CNN2026.03 | 90.6 | |
| EdgeFLowSeqCluster size Nm=10, Local epoch K=5, Mini-batch size=64, Backbone=6-layer CNN2026.03 | 90.53 | |
| EdgeFLowRandCluster size Nm=10, Local epoch K=5, Mini-batch size=64, Backbone=6-layer CNN2026.03 | 90.13 | |
| EdgeFD2025.08 | 89.9 | |
| FedAvgN (Number of clients)=10, R (Communication rounds)=302026.05 | 89.38 | |
| IndLearn2025.08 | 88.8 | |
| FedRepN (Number of clients)=10, R (Communication rounds)=302026.05 | 88.22 | |
| FibFL++N (Number of clients)=10, R (Communication rounds)=302026.05 | 87.88 | |
| FedMD2025.08 | 87.25 | |
| Selective-FD2025.08 | 87.16 | |
| FedED2025.08 | 86.88 | |
| DS-FL2025.08 | 86.62 | |
| PLS2025.08 | 86.52 | |
| FKD2025.08 | 86.14 | |
| RDFLN (Number of clients)=10, R (Communication rounds)=302026.05 | 85.11 | |
| FibFL+N (Number of clients)=10, R (Communication rounds)=302026.05 | 81.34 | |
| FibFLN (Number of clients)=10, R (Communication rounds)=302026.05 | 80.47 |