Table 5 5-Fold cross-validation performances of methods on stringent Chen multispecies datasets
Similarity | Methods | Accuracy (%) | F1-Score (%) |
|---|---|---|---|
Any | PIPR (2019) | 98.19 | 98.17 |
DeepTrio (2022) | 98.20 | 98.20 | |
TAGPPI (2022) | 99.15 | 99.15 | |
Our xCAPT5 | 99.72 | 99.61 | |
\(\le 40\%\) | PIPR (2019) | 98.29 | 98.28 |
DeepTrio (2022) | 97.83 | 97.98 | |
TAGPPI (2022) | 99.10 | 99.16 | |
Our xCAPT5 | 99.76 | 99.60 | |
\(\le 25\%\) | PIPR (2019) | 97.91 | 98.08 |
DeepTrio (2022) | 97.52 | 97.75 | |
TAGPPI (2022) | 98.99 | 99.06 | |
Our xCAPT5 | 99.74 | 99.61 | |
\(\le 10\%\) | PIPR (2019) | 97.54 | 97.79 |
DeepTrio (2022) | 97.32 | 97.62 | |
TAGPPI (2022) | 98.97 | 99.08 | |
Our xCAPT5 | 99.70 | 99.53 | |
\(\le 1\%\) | PIPR (2019) | 97.51 | 97.80 |
DeepTrio (2022) | 97.11 | 97.47 | |
TAGPPI (2022) | 98.89 | 98.89 | |
xCAPT5 | 99.73 | 99.60 |