Table 3 Performance of stage-ethnicity extraction (macro average)
From: Weakly supervised learning of biomedical information extraction from curated data
Method | Precision | Recall | F1 Score |
|---|---|---|---|
Baseline | 0.5972 | 1.0000 | 0.7478 |
Cost-insensitive | 0.7408 | 0.7943 | 0.7666 |
Cost-sensitive | 0.7893 | 0.8757 | 0.8302 |