Table 1 Result of 10-fold cross-validation on rf452 using several classifiers
From: StackDPP: a stacking ensemble based DNA-binding protein prediction model
Classifier | ACC (%) | SN (%) | SP (%) | MCC |
|---|---|---|---|---|
Decision tree | 84.27 | 85.53 | 83.00 | 0.6874 |
Logistic Regression | 88.35 | 89.75 | 86.94 | 0.7682 |
Random Forest | 90.03 | 88.76 | 91.30 | 0.8025 |
SVC (RBF) | 91.01 | 90.45 | 91.57 | 0.8213 |
SVC (RBF, tuned) | 91.96 | 91.72 | 92.28 | 0.8412 |
SVC (linear) | 86.03 | 87.64 | 84.41 | 0.7224 |
SVC (polynomial) | 89.46 | 93.68 | 85.25 | 0.7935 |
SVC (sigmoid) | 88.83 | 89.61 | 88.06 | 0.7782 |
Extra Tree | 90.81 | 90.31 | 91.29 | 0.8167 |
Gaussian Naive Bayes | 88.76 | 89.74 | 87.78 | 0.7772 |
Adaboost | 87.22 | 88.61 | 85.82 | 0.7462 |
Linear Discriminant Analysis | 87.36 | 89.32 | 85.39 | 0.7484 |
K-nearest neighbour | 89.39 | 91.01 | 87.78 | 0.7895 |
Bagging classifier | 89.25 | 86.93 | 91.58 | 0.7873 |
Bagging with SVC (RBF) | 91.08 | 90.73 | 91.43 | 0.8226 |