Anomaly Detection on Mammography (FDR/Power Reporting)
1.7False Discovery Rate (FDR)Weighted EDF
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Weighted EDFInference Strategy=Deterministic, Homogeneous pruning=true, n_train (calibration size)=5,461, n_test (test size)=2,7962026.03 | 1.7 | 3.8 | |
| Weighted EDFInference Strategy=Randomized, Homogeneous pruning=true, n_train (calibration size)=5,461, n_test (test size)=2,7962026.03 | 1.7 | 4.3 | |
| EDFInference Strategy=Deterministic, n_train (calibration size)=5,461, n_test (test size)=2,7962026.03 | 1.9 | 5.2 | |
| EDFInference Strategy=Randomized, n_train (calibration size)=5,461, n_test (test size)=2,7962026.03 | 2.6 | 6.9 | |
| Weighted KDEInference Strategy=Deterministic, Homogeneous pruning=true, n_train (calibration size)=5,461, n_test (test size)=2,7962026.03 | 3.7 | 7.7 | |
| KDEInference Strategy=Deterministic, n_train (calibration size)=5,461, n_test (test size)=2,7962026.03 | 4.5 | 8.5 |