Table 9 Performance evaluation of different feature selection and classification algorithm combinations on mathematics dataset.
Combination | Feature Selection | Classification | Accuracy | Precision | Recall | F1-Score |
|---|---|---|---|---|---|---|
1 | Chi-square | DT | 90.94 | 91.45 | 92.34 | 91.11 |
2 | Chi-square | RF | 95.80 | 92.39 | 88.12 | 90.05 |
3 | Chi-square | SVM | 88.76 | 93.12 | 89.12 | 91.15 |
4 | Chi-square | NN | 89.49 | 90.45 | 91.45 | 91.15 |
5 | Chi-square | NB | 90.57 | 92.56 | 92.00 | 91.15 |
6 | Chi-square | J48 | 90.59 | 94.42 | 90.05 | 92.45 |
7 | Information gain | DT | 94.17 | 89.90 | 93.47 | 91.15 |
8 | Information gain | RF | 91.09 | 92.88 | 86.66 | 89.09 |
9 | Information gain | SVM | 89.85 | 94.78 | 90.69 | 92.67 |
10 | Information gain | NN | 89.66 | 88.12 | 88.35 | 88.67 |
11 | Information gain | NB | 88.76 | 96.47 | 91.23 | 93.03 |
12 | Information gain | J48 | 90.57 | 91.45 | 88.20 | 89.09 |
13 | Correlation heat map | DT | 92.02 | 91.67 | 90.00 | 91.69 |
14 | Correlation heat map | RF | 93.46 | 91.67 | 89.99 | 87.34 |
15 | Correlation heat map | SVM | 91.30 | 90.43 | 91.10 | 89.47 |
16 | Correlation heat map | NN | 92.39 | 92.09 | 88.37 | 89.73 |
17 | Correlation heat map | NB | 92.02 | 91.90 | 91.13 | 91.76 |
18 | Correlation heat map | J48 | 94.20 | 93.56 | 92.05 | 92.05 |
