Table 2.
Performance comparison of our doc2vec + RF model with DeNovo and Alguwzizani et al.’s method using the test set of DeNovo.
| Method | SN (%) | SP (%) | ACC (%) | PPV (%) | NPV (%) | MCC | AUC | F1 (%) |
|---|---|---|---|---|---|---|---|---|
| Our model | 90.33 | 96.17 | 93.23 | 95.99 | 90.74 | 0.866 | 0.981 | 93.07 |
| Alguwzizani et al.’s SVMa,b | 86.35 | 86.59 | 86.47 | 86.56 | 86.39 | 0.729 | 0.926 | NA |
| DeNovob,c | 80.71 | 83.06 | 81.90 | NA | NA | NA | NA | NA |
a
The corresponding values were retrieved from [54].
b
NA means the corresponding parameter is not available.
c
The corresponding values were retrieved from [43]. SN: Sensitivity, SP: Specificity, ACC: Accuracy, PPV: Positive Predictive Value (PPV = Precision); NPV: Negative Predictive Value (NPV = TN/(TN + FN)); MCC: Matthews Correlation Coefficient; AUC: the area under the ROC curve; F1 = 2 × (Precision × Recall)/(Precision + Recall).