Skip to main content
. 2021 Sep 28;7:e715. doi: 10.7717/peerj-cs.715

Table 9. Results of the proposed model with TL.

Model Accuracy (%) Specificity (%) Recall (%) Precision (%) F1-Score (%)
From Scratch 97.6 96.7 98.5 96.7 97.6
With TL 99.5 99.1 99.9 99.0 99.4
Unseen Test Set 93.7 89.4 100 86.6 92.8