Table 2.
Accuracy values for the training and test data. Best accuracies are highlighted in bold
| 20-fold CV for training data | |||
|---|---|---|---|
| Classifier | Average accuracy (%) | Standard deviation | Accuracy for test data (%) |
| RidgeClassifier | 88.4 | 4.8 | 88.4 |
| SGDClassifier | 88.2 | 5.1 | 88.6 |
| PassiveAggressiveClassifier | 88.0 | 4.7 | 89.3 |
| LogisticRegression | 87.4 | 5.6 | 86.6 |
| MultinomialNB | 82.4 | 7.9 | 79.5 |
| ComplementNB | 84.1 | 7.0 | 82.0 |
| BernoulliNB | 82.1 | 7.6 | 83.7 |
| DecisionTreeClassifier | 86.4 | 5.9 | 92.7 |
| RandomForestClassifier | 87.4 | 6.0 | 88.0 |
| BaggingClassifier | 87.5 | 5.1 | 95.3 |
| KNeighborsClassifier | 82.1 | 5.4 | 81.1 |
| AdaBoostClassifier | 84.0 | 6.1 | 91.1 |
| SVC | 88.0 | 4.5 | 89.1 |
| MLPClassifier | 87.1 | 5.7 | 88.2 |