Table 1.
Performances of the models trained by RF.
Metrics | Sensitivity | Specificity | Accuracy | Precision | Misclassification (%) | OOB error |
---|---|---|---|---|---|---|
1195-Rf-acc | 0.81 | 0.65 | 0.73 | 0.71 | 27 | 0.261 |
200-Rf-acc | 0.86 | 0.82 | 0.84 | 0.84 | 16 | 0.154 |
48-Rf-acc | 0.93 | 0.80 | 0.87 | 0.83 | 13 | 0.131 |
40-Rf-acc | 0.85 | 0.88 | 0.87 | 0.87 | 13 | 0.131 |
30-Rf-acc | 0.83 | 0.90 | 0.87 | 0.90 | 13 | 0.131 |
20-Rf-acc | 0.90 | 0.85 | 0.88 | 0.86 | 12 | 0.119 |
10-Rf-acc | 0.85 | 0.86 | 0.85 | 0.85 | 15 | 0.142 |
5-Rf-acc | 0.86 | 0.85 | 0.85 | 0.86 | 14 | 0.142 |
Table of performance of 9 models trained by RF on different datasets. Six common evaluation metrics were used.