Tabular Synthetic Data Generation on Breast Cancer low-order real-world (test)
97.68AccuracyUB
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| UBepsilon=1.0, Description=Upper bound performance trained on real data2026.06 | 97.68 | 97.56 | 0.078 | 0.13 | 0.175 | 97.48 | 95.64 | |
| AIMepsilon=1.02026.06 | 89.25 | 87.82 | 0.198 | 0.332 | 0.441 | 68.17 | 12.65 | |
| Tab-PEepsilon=1.02026.06 | 88.48 | 87.01 | 0.162 | 0.272 | 0.366 | 69.02 | 15.24 | |
| RAP++epsilon=1.02026.06 | 84.81 | 83.97 | 0.227 | 0.369 | 0.483 | 64.62 | 9.63 | |
| PrivMRFepsilon=1.02026.06 | 60.41 | 37.63 | 0.18 | 0.306 | 0.41 | 51.01 | 16.83 | |
| PrivGSDepsilon=1.02026.06 | 60.02 | 37.49 | 0.235 | 0.379 | 0.493 | 60.3 | 21.27 | |
| PrivSynepsilon=1.02026.06 | 51.6 | 38.99 | 0.216 | 0.353 | 0.464 | 60.39 | 21.78 | |
| GEMepsilon=1.02026.06 | 50.74 | 44.91 | 0.181 | 0.306 | 0.409 | 69.6 | 28.64 |