Table 6 Discrimination of channels and pores using different machine learning approaches
From: Functional discrimination of membrane proteins using machine learning techniques
Method | 5-fold cross-validation | ||||
|---|---|---|---|---|---|
Sensitivity (%) | Specificity (%) | F-measure | Accuracy (%) | ||
Channel | Pore | ||||
Bayesnet | 94.1 | 81.4 | 0.910 | 0.857 | 88.9 |
Naive Bayes | 92.5 | 88.4 | 0.923 | 0.887 | 90.8 |
Logistic function | 92.0 | 89.1 | 0.922 | 0.888 | 90.8 |
Neural network | 93.0 | 91.5 | 0.935 | 0.915 | 92.4 |
RBF network | 92.5 | 88.4 | 0.923 | 0.887 | 90.8 |
Support vector machines | 95.2 | 88.4 | 0.937 | 0.905 | 92.4 |
k-nearest neighbor | 89.8 | 86.8 | 0.903 | 0.862 | 88.6 |
Bagging meta learning | 89.8 | 83.7 | 0.894 | 0.844 | 87.3 |
Classification via Regression | 88.2 | 85.3 | 0.889 | 0.843 | 87.0 |
Decision tree J4.8 | 86.1 | 78.3 | 0.856 | 0.789 | 82.9 |
NBTree | 90.9 | 83.7 | 0.899 | 0.850 | 88.0 |
Partial decision tree | 87.2 | 79.1 | 0.865 | 0.800 | 83.9 |