Table 5.
The 5-fold cross-validation test results obtained from different classification algorithms with the same feature extraction method on benchmark data-set only containing splice donor site sequences.
| Predictor | ACC(%) | MCC(%) | Sn(%) | Sp(%) |
|---|---|---|---|---|
| iSS-PCa | 90.56 | 81.13 | 90.09 | 91.04 |
| iSS-SVMb | 77.59 | 55.25 | 75.68 | 79.50 |
| iSS-RFc | 83.13 | 66.38 | 80.11 | 86.14 |
| iSS-libD3Cd | 83.38 | 67.09 | 78.43 | 88.32 |
aThe predictor with SAE proposed in this paper.
bThe predictor with SVM created in WEKA with the default parameters.
cThe predictor with Random Forest (RF) created in WEKA.
dThe predictor with an ensemble classifier libD3C.