Image Classification on MNIST (Accuracy, Energy, and Speed)
99AccuracyFedRank
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| FedRankData Distribution=IID, Model Architecture=LeNet5, Training Termination=Reached target accuracy (99%)2026.06 | 99 | 40.1 | 1.56 | |
| FedGCSData Distribution=IID, Model Architecture=LeNet5, Training Termination=Reached target accuracy (99%)2026.06 | 99 | — | — | |
| EvoCSFLData Distribution=IID, Model Architecture=LeNet5, Training Termination=Reached target accuracy (99%)2026.06 | 99 | 41.4 | 2.28 | |
| OortData Distribution=IID, Model Architecture=LeNet52026.06 | 98.97 | 44.2 | 2.12 | |
| TiFLData Distribution=IID, Model Architecture=LeNet52026.06 | 98.86 | 60 | 1.72 | |
| AFLData Distribution=IID, Model Architecture=LeNet52026.06 | 98.8 | 61.3 | 1.62 | |
| FedAvgData Distribution=IID, Model Architecture=LeNet52026.06 | 98.75 | 100 | 1 | |
| EvoCSFLData Distribution=Non-IID (δ=0.01), Model Architecture=LeNet52026.06 | 96.52 | 45.5 | 1.92 | |
| FedGCSData Distribution=Non-IID (δ=0.01), Model Architecture=LeNet52026.06 | 95.83 | — | — | |
| FedRankData Distribution=Non-IID (δ=0.01), Model Architecture=LeNet52026.06 | 93.67 | 47.4 | 1.48 | |
| OortData Distribution=Non-IID (δ=0.01), Model Architecture=LeNet52026.06 | 93.18 | 51.4 | 1.84 | |
| TiFLData Distribution=Non-IID (δ=0.01), Model Architecture=LeNet52026.06 | 91.59 | 58.7 | 1.77 | |
| AFLData Distribution=Non-IID (δ=0.01), Model Architecture=LeNet52026.06 | 90.23 | 63.5 | 1.53 | |
| FedAvgData Distribution=Non-IID (δ=0.01), Model Architecture=LeNet52026.06 | 87.61 | 100 | 1 |