Image Classification on MNIST Cross-silo
95.49Classification AccuracyFLTrust
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| FLTrustAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 95.49 | 6.52 | 7.35 | 8.2 | 8.25 | — | 7.4 | 4.98 | 7.2 | |
| SgnGAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 95.31 | 24.57 | 81.18 | 12.66 | 25.21 | — | 5.3 | 3.99 | 6.87 | |
| FedAvgAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 95.02 | 29.42 | 81.43 | 9.39 | 17.7 | 85.22 | 4.98 | 84.92 | 83.62 | |
| FoolsGoldAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 94.8 | 35.91 | 36.11 | 4.48 | 3.56 | — | 2.49 | 4.36 | 3.38 | |
| Tr-MeanAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 94.69 | 14.38 | 15.19 | 10.97 | 13.64 | 21.27 | 6.91 | 5.55 | 14 | |
| MedianAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 94.61 | 13.73 | 12.21 | 12.5 | 11.43 | 14.31 | 8 | 5.15 | 8.28 | |
| M-KrumAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 94.57 | 9.34 | 12.5 | 10.32 | 9.08 | 11.91 | 7.68 | 5.58 | 4.69 | |
| FreqFedAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 94.32 | 5.95 | 80.87 | 16.14 | 35.18 | — | 9.69 | 5.72 | 6.79 | |
| CCAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 94.08 | 13.13 | 80.1 | 13.49 | 23.57 | — | 4.62 | 3.32 | 5.6 | |
| FLAMEAggregation Setting=Cross-silo, Number of FL clients=100, Global iterations=275, Model architecture=4 layer DNN model2026.04 | 93.33 | 3.38 | 79.24 | 14.93 | 34.07 | — | 11.34 | 3.44 | 3.56 |