Table 4.
Performance comparison of SVM on different feature sets for binary experiments. Bold font indicates the best result obtained against each feature set.
| Feature Set | Classifier | Accuracy (%) | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Error (%) |
|---|---|---|---|---|---|---|---|
| MFCC | SVM-L | 89.88 | 76.19 | 94.44 | 82.05 | 92.25 | 10.12 |
| SVM-Q | 89.29 | 80.95 | 92.06 | 77.27 | 93.55 | 10.71 | |
| SVM-C | 92.26 | 88.1 | 93.65 | 82.22 | 95.93 | 7.74 | |
| SVM-G | 75.6 | 7.14 | 98.41 | 60 | 76.07 | 24.4 | |
| 1D-LTP | SVM-L | 94.05 | 88.1 | 96.03 | 88.1 | 96.03 | 5.95 |
| SVM-Q | 94.05 | 83.33 | 97.62 | 92.11 | 94.62 | 5.95 | |
| SVM-C | 91.07 | 76.19 | 96.03 | 86.49 | 92.37 | 8.93 | |
| SVM-G | 86.31 | 47.62 | 99.21 | 95.24 | 85.03 | 13.69 | |
| MFCC+1D-LTP | SVM-L | 94.05 | 90.48 | 95.24 | 86.36 | 96.77 | 5.95 |
| SVM-Q | 94.05 | 88.1 | 96.03 | 88.1 | 96.03 | 5.95 | |
| SVM-C | 95.83 | 92.86 | 96.83 | 90.7 | 97.6 | 4.17 | |
| SVM-G | 93.45 | 88.1 | 95.24 | 86.05 | 96 | 6.55 |