Table 3 Feature selection model performance comparison.
Feature selection model | Accuracy (%) | Precision (%) | Recall (%) | F1-score (%) |
|---|---|---|---|---|
No feature selection | 92.45 | 91.23 | 90.88 | 91.05 |
Principal component analysis (PCA) | 93.1 | 91.95 | 91.22 | 91.58 |
Recursive feature elimination (RFE) | 94.02 | 92.86 | 92.1 | 92.47 |
Mutual information | 93.78 | 92.34 | 91.87 | 92.1 |
Chi-square test | 93.45 | 92.05 | 91.56 | 91.8 |
Proposed (correlation-based feature selection - CFS) | 98.39 | 97.82 | 98.02 | 97.91 |
L1 regularization (Lasso) | 93.68 | 92.28 | 91.82 | 92.05 |
