Table 2 The performance of hybrid module, which combines HMM and SVM model (using NT25+R).
From: Prediction of nuclear proteins using SVM and HMM models
SVM Threshold | Sensitivity | Specificity | Accuracy | MCC |
|---|---|---|---|---|
-1.0 | 99.19 | 88.03 | 90.95 | 0.80 |
-0.9 | 99.04 | 88.86 | 91.53 | 0.82 |
-0.8 | 98.86 | 89.61 | 92.03 | 0.82 |
-0.7 | 98.63 | 90.37 | 92.53 | 0.83 |
-0.6 | 98.34 | 91.08 | 92.98 | 0.84 |
-0.5 | 97.82 | 91.67 | 93.28 | 0.85 |
-0.4 | 97.49 | 92.39 | 93.72 | 0.85 |
-0.3 | 96.86 | 92.96 | 93.98 | 0.86 |
-0.2 | 96.13 | 93.41 | 94.12 | 0.86 |
-0.1 | 95.46 | 94.05 | 94.42 | 0.86 |
0.0 | 94.80 | 94.55 | 94.61 | 0.87 |
0.1 | 94.02 | 95.00 | 94.75 | 0.87 |
0.2 | 92.80 | 95.22 | 94.59 | 0.86 |
0.3 | 92.07 | 95.77 | 94.80 | 0.87 |
0.4 | 90.70 | 96.18 | 94.75 | 0.87 |
0.5 | 89.52 | 96.62 | 94.77 | 0.87 |
0.6 | 87.90 | 96.91 | 94.55 | 0.86 |
0.7 | 86.27 | 97.25 | 94.38 | 0.85 |
0.8 | 85.21 | 97.57 | 94.34 | 0.85 |
0.9 | 83.58 | 97.70 | 94.01 | 0.84 |
1.0 | 82.66 | 98.00 | 93.99 | 0.84 |