Table 4 Comparative analysis based on dataset feature using proposed and existing technique.
Classifiers | Feature class | Parameters | |||
Prediction accuracy | Mean average precision (MAP) | F-1 score | Latency | ||
CNN | Demographic features | 68 | 72 | 63 | 71 |
Academic background features | 73 | 76 | 67 | 74 | |
Behavioral factors | 75 | 79 | 71 | 77 | |
KNN | Demographic features | 75 | 82 | 74 | 67 |
Academic background features | 77 | 84 | 77 | 73 | |
Behavioral factors | 80 | 87 | 80 | 77 | |
SVM | Demographic features | 82 | 80 | 83 | 80 |
Academic background features | 85 | 83 | 85 | 82 | |
Behavioral factors | 87 | 88 | 89 | 85 | |
Random forest | Demographic features | 81 | 82 | 84 | 83 |
Academic background features | 83 | 84 | 88 | 85 | |
Behavioral factors | 85 | 86 | 90 | 87 | |
Proposed B-FCTNN_SRSO model | Demographic features | 89 | 88 | 93 | 92 |
Academic background features | 94 | 91 | 95 | 94 | |
Behavioral factors | 98 | 95 | 97 | 96 |