Skip to main content
. 2021 Apr 12;18(8):4069. doi: 10.3390/ijerph18084069

Table 4.

Performance among Bi-LSTM, BERT, SVM, and NB on the test set.

Accuracy Precision Recall F1 Score AUC
Bi-LSTM-128, learning rate = 0.0001, epoch = 50 89.8% 44.0% 47.2% 45.5% 85.8%
BERT, learning rate = 0.0001, epoch = 3 91.6% 93.4% 97.6% 95.5% 84.7%
SVM-linear 92.3% 19.5% 78.6% 31.2% 85.6%
Complement NB 88.8% 23.0% 32.8% 27.1% 62.7%