Research Article
SABinder: A Web Service for Predicting Streptavidin-Binding Peptides
Table 3
The prediction performances of various machine learning methods.
| Machine learning methods | Sn (%) (mean ± std) | Sp (%) (mean ± std) | Acc (%) (mean ± std) | MCC (mean ± std) |
| Support vector machine | 84.72 ± 2.19 | 93.67 ± 1.87 | 89.20 ± 1.23 | 0.79 ± 0.02 | Naïve Bayes | 78.85 ± 3.90 | 77.40 ± 1.73 | 78.11 ± 2.47 | 0.56 ± 0.05 | Random Forest | 84.80 ± 2.30 | 88.00 ± 5.22 | 86.41 ± 2.41 | 0.73 ± 0.05 | Decision Tree J48 | 76.90 ± 1.10 | 88.24 ± 4.31 | 82.57 ± 2.00 | 0.66 ± 0.04 | RBF network | 79.00 ± 4.18 | 78.50 ± 2.33 | 78.74 ± 2.57 | 0.58 ± 0.05 | Logistic Function | 76.40 ± 3.22 | 67.83 ± 3.82 | 72.11 ± 3.23 | 0.44 ± 0.06 |
|
|
Std: standard deviation.
|