TABLE II. Comparison of Different Missing Data Filling Strategy on Random Forest, Gradient Boosting, MLP and Our IE-Net, in Terms of Accuracy, Recall, Precision and AUC (%).
4]*Methods | 4]*Fill | Evaluation Metrics (mean std) | |||
---|---|---|---|---|---|
ACC | Recall | Precision | AUC | ||
2]*GradientBoosting | Zeros | 85.403.08 | 79.725.54 | 89.314.06 | 85.223.36 |
Average | 85.943.86 | 78.855.14 | 89.215.13 | 84.962.02 | |
Median | 84.763.15 | 78.825.87 | 89.144.09 | 84.663.40 | |
2]*Random Forest | Zeros | 84.763.65 | 80.626.77 | 85.216.07 | 84.424.19 |
Average | 82.214.83 | 79.814.39 | 84.627.10 | 83.832.82 | |
Median | 84.443.66 | 79.996.967 | 85.465.68 | 84.104.18 | |
2]*MLP | Zeros | 82.062.99 | 74.985.29 | 88.035.25 | 82.193.55 |
Average | 55.715.30 | 0.000.00 | 0.000.00 | - | |
Median | 55.715.30 | 0.000.00 | 0.000.00 | - | |
2]*IE-Net | Zeros | 64.419.74 | 92.352.36 | 83.0511.65 | 71.706.60 |
Average | 63.8110.65 | 91.813.32 | 82.6510.65 | 71.476.81 | |
Median | 63.459.41 | 92.512.15 | 83.4610.71 | 71.976.16 |