Skip to main content
. 2021 Feb 18;25(5):1347–1357. doi: 10.1109/JBHI.2021.3060035

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 Inline graphic std)
ACC Recall Precision AUC
2]*GradientBoosting Zeros 85.40Inline graphic3.08 79.72Inline graphic5.54 89.31Inline graphic4.06 85.22Inline graphic3.36
Average 85.94Inline graphic3.86 78.85Inline graphic5.14 89.21Inline graphic5.13 84.96Inline graphic2.02
Median 84.76Inline graphic3.15 78.82Inline graphic5.87 89.14Inline graphic4.09 84.66Inline graphic3.40
2]*Random Forest Zeros 84.76Inline graphic3.65 80.62Inline graphic6.77 85.21Inline graphic6.07 84.42Inline graphic4.19
Average 82.21Inline graphic4.83 79.81Inline graphic4.39 84.62Inline graphic7.10 83.83Inline graphic2.82
Median 84.44Inline graphic3.66 79.99Inline graphic6.967 85.46Inline graphic5.68 84.10Inline graphic4.18
2]*MLP Zeros 82.06Inline graphic2.99 74.98Inline graphic5.29 88.03Inline graphic5.25 82.19Inline graphic3.55
Average 55.71Inline graphic5.30 0.00Inline graphic0.00 0.00Inline graphic0.00 -
Median 55.71Inline graphic5.30 0.00Inline graphic0.00 0.00Inline graphic0.00 -
2]*IE-Net Zeros 64.41Inline graphic9.74 92.35Inline graphic2.36 83.05Inline graphic11.65 71.70Inline graphic6.60
Average 63.81Inline graphic10.65 91.81Inline graphic3.32 82.65Inline graphic10.65 71.47Inline graphic6.81
Median 63.45Inline graphic9.41 92.51Inline graphic2.15 83.46Inline graphic10.71 71.97Inline graphic6.16