VulDistilBERT: A CPS Vulnerability Severity Prediction Method Based on Distillation Model

Table 10

Results of comparing the performance of different classifiers on a data-augmented dataset.

Models	Classifiers	Best epoch	Loss	Accuracy	Precision	Recall	F1 scores
Pretrained DistilBERT	Linear	32	0.927	0.65	0.66	0.65	0.65
Fine-tuned DistilBERT	Linear	3	0.221	0.92	0.92	0.92	0.92
Fine-tuned DistilBERT	CNN	4	0.225	0.91	0.92	0.92	0.92
Fine-tuned DistilBERT	LSTM	3	0.228	0.91	0.92	0.92	0.92