Table 4.
The comparison of our study with state-of-the-art methods.
| Study | Group and No. of Data | Classifier | Acc | Sen | Spe | AUC |
|---|---|---|---|---|---|---|
| Ours | PE (n = 143)-TE (n = 143) | AdaBoost | 95% | 92% | 97% | 0.99 |
| PL (n = 119)-TL (n = 119) | 93% | 90% | 93% | 0.98 | ||
| [42] | PE (n = 135)-TE (n = 143) | RF | 92% | 88% | 96% | 0.88 |
| PL (n = 111)-TL (n = 119) | 93% | 89% | 97% | 0.80 | ||
| [43] | PE (n = 93)-TE (n = 93) | QDA | 97% | 100% | 95% | N.A |
| PL (n = 57)-TL (n = 57) | 100% | 100% | 100% | N.A | ||
| [35] | PE (n = 140)-TE (n = 143) | QDA | 100% | 100% | 100% | 1.0 |
| [44] | PE-TE, n is not reported. | SVM | 96.5% | 94% | 99% | 0.99 |
| PL-TL, n is not reported. | 92.5% | 88% | 97% | 0.98 |