Federated Learning Image Classification on MNIST (test)
84AccuracyFLANDERS + FedAvg
Evaluation results
| Method | Links | |
|---|---|---|
| FLANDERS + FedAvgAttack Strategy=LIE, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 84 | |
| FLANDERS + FedAvgAttack Strategy=OPT, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 84 | |
| FLANDERS + FedAvgAttack Strategy=AGR-MM, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 82 | |
| FLANDERS + FedAvgAttack Strategy=GAUSS, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 75 | |
| FLDetector + FedAvgAttack Strategy=OPT, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 23 | |
| FedAvgAttack Strategy=OPT, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 21 | |
| FedAvgAttack Strategy=GAUSS, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 18 | |
| FLDetector + FedAvgAttack Strategy=GAUSS, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 15 | |
| FLDetector + FedAvgAttack Strategy=LIE, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 13 | |
| FLDetector + FedAvgAttack Strategy=AGR-MM, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 12 | |
| FedAvgAttack Strategy=LIE, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 11 | |
| FedAvgAttack Strategy=AGR-MM, Poisoning Ratio (r)=0.8, Number of Rounds (T)=502023.03 | 11 |