Semi-supervised Classification on MedMNIST Breast (10% train 90% test)
85.7AccuracyARMA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ARMALoss Function=MERIT + Cut, lambda_con=0.00012026.05 | 85.7 | 89.2 | 91.6 | 90.3 | |
| ARMA-C3Loss Function=-, lambda_con=0.00012026.05 | 85.5 | 89 | 91.7 | 90.2 | |
| ARMALoss Function=Baseline, lambda_con=N/A2026.05 | 84.6 | 87.6 | 92 | 89.7 | |
| ARMALoss Function=Cut, lambda_con=0.0012026.05 | 84.4 | 87.7 | 91.6 | 89.6 | |
| ARMALoss Function=Mod, lambda_con=0.0012026.05 | 84.3 | 87.3 | 92.1 | 89.6 | |
| APPNPLoss Function=-, lambda_con=N/A2026.05 | 84.1 | 86.7 | 92.6 | 89.5 | |
| SVCLoss Function=-, lambda_con=-2026.05 | 83.9 | 83.6 | 97.2 | 89.9 | |
| MLPClassifierLoss Function=-, lambda_con=-2026.05 | 83 | 89.7 | 86.8 | 88.2 | |
| JacobiConvLoss Function=-, lambda_con=N/A2026.05 | 81.7 | 86.2 | 90 | 87.7 | |
| XGBoostLoss Function=-, lambda_con=-2026.05 | 80.9 | 82.5 | 93.7 | 87.7 | |
| Random ForestLoss Function=-, lambda_con=-2026.05 | 79.8 | 79.1 | 98.5 | 87.7 | |
| ARMALoss Function=DGI + Mod, lambda_con=0.012026.05 | 79 | 89.7 | 80.6 | 84.8 | |
| GATLoss Function=Mod, lambda_con=0.0012026.05 | 75.4 | 86.6 | 78.8 | 82.3 | |
| GATLoss Function=MERIT + Cut, lambda_con=0.00012026.05 | 75.3 | 89.8 | 75 | 81.2 | |
| GATLoss Function=MERIT + Mod, lambda_con=0.00012026.05 | 75.1 | 89.9 | 74.4 | 81 | |
| GATLoss Function=Baseline, lambda_con=N/A2026.05 | 74 | 85.8 | 77.7 | 80.9 | |
| ARMALoss Function=DGI + Cut, lambda_con=0.012026.05 | 73 | 90.8 | 70.4 | 79 | |
| GATLoss Function=Cut, lambda_con=0.00012026.05 | 72.5 | 87.7 | 73.4 | 78.8 | |
| GATLoss Function=DGI + Mod, lambda_con=0.012026.05 | 69.5 | 88.8 | 66.8 | 75.8 | |
| GATLoss Function=DGI + Cut, lambda_con=0.0012026.05 | 67.7 | 87.6 | 65.4 | 74.2 |