Table 2 Comparison of performance with different dimensions of hidden states for sub-classifiers \(f_{5}, f_{10}, f_{15}\)
Dimension | Accuracy (%) | Precision (%) | Sensitivity (%) | Specificity (%) | \(F_{1}\)Score (%) |
|---|---|---|---|---|---|
Sub-classifier \(f_{5}\) | |||||
10 | 85.98 | 92.87 | 88.78 | 66.04 | 90.78 |
25 | 88.53 | 92.60 | 92.70 | 74.36 | 92.63 |
50 | 90.35 | 93.25 | 94.42 | 79.61 | 93.83 |
75 | 90.18 | 94.21 | 93.10 | 76.84 | 93.65 |
Sub-classifier \(f_{10}\) | |||||
10 | 79.90 | 82.59 | 77.56 | 77.35 | 80.00 |
25 | 85.22 | 84.22 | 87.95 | 86.40 | 86.05 |
50 | 89.15 | 89.24 | 89.90 | 89.05 | 89.57 |
75 | 88.59 | 88.02 | 90.27 | 89.24 | 89.13 |
Sub-classifier \(f_{15}\) | |||||
10 | 90.67 | 87.68 | 74.42 | 91.51 | 80.51 |
25 | 94.53 | 91.11 | 87.39 | 95.66 | 89.21 |
50 | 96.17 | 93.03 | 92.10 | 97.25 | 92.57 |
75 | 96.12 | 92.96 | 91.99 | 97.21 | 92.47 |