Skip to main content
. 2024 Sep 12;2024:baae095. doi: 10.1093/database/baae095

Table 2.

Performance of the best experiment on the RegulaTome development set

Precision Recall F1-score
Model-1 69.1 67.4 68.3
Model-2 68.3 66.3 67.3
Model-3 69.7 66.8 68.2
Model-4 68.6 67.3 67.9
Average 68.9 67.0 67.9
SD 0.61 0.51 0.45

The best model (highlighted in bold) is used to perform a run on the held-out test set and for a large-scale run on the entire biomedical scientific literature.