Table 4 F1-score performance for both the models and ensemble across all the subclasses
Label | F1-score (%) | |||||
|---|---|---|---|---|---|---|
RoBERTa base | RoBERTa large | BioBERT | PubMedBERT | COVID-Twitter | Ensemble | |
EPI | 88.17 | 88.05 | 88.38 | 88.70 | 87.26 | 89.47 |
BASIC | 78.15 | 78.85 | 79.20 | 78.13 | 78.36 | 80.47 |
OTHER | 78.44 | 79.22 | 79.86 | 80.72 | 76.71 | 81.97a |
micro avg | 84.01 | 84.26 | 84.68 | 84.99 | 82.99 | 86.10a |
macro avg | 81.59 | 82.04 | 82.48 | 82.51 | 80.77 | 83.97a |