Table 2.
The classification results using the combination of entropy features for the University of Bonn dataset.
| Feature | Accuracy (%) | Sensitivity (%) | Specificity (%) |
|---|---|---|---|
| Single feature | |||
| ApEn | 62.60 | 42 | 83.20 |
| SampEn | 64.15 | 56.30 | 72 |
| PermEn | 59.80 | 62.20 | 57.40 |
| FuzzyEn | 64.85 | 48.90 | 80.80 |
| LogEn | 66.90 | 64.70 | 69.10 |
| Two features | |||
| SampEn, ApEn | 72.60 | 64.80 | 80.40 |
| SampEn, LogEn | 79.40 | 72.60 | 86.2 |
| SampEn, PermEn | 67.35 | 62.80 | 71.90 |
| SampEn, FuzzyEn | 75.80 | 66.80 | 84.80 |
| ApEn, LogEn | 79.75 | 72 | 87.5 |
| ApEn, PermEn | 68.70 | 56.90 | 80.50 |
| ApEn, FuzzyEn | 78.05 | 70.60 | 85.50 |
| LogEn, PermEn | 76.45 | 77.40 | 75.50 |
| LogEn, FuzzyEn | 79.60 | 71.20 | 88 |
| PermEn, FuzzyEn | 72.85 | 70.40 | 75.30 |
| Three features | |||
| SampEn, ApEn, LogEn | 84.35 | 79.20 | 89.50 |
| SampEn, ApEn, PermEn | 73.50 | 70.40 | 76.60 |
| SampEn, ApEn, FuzzyEn | 82.35 | 77.40 | 87.30 |
| SampEn, PermEn, LogEn | 82.75 | 77.70 | 87.80 |
| SampEn, LogEn, FuzzyEn | 87.35 | 86.10 | 88.60 |
| SampEn, FuzzyEn, PermEn | 84.15 | 78.50 | 89.80 |
| ApEn, LogEn, PermEn | 86.70 | 81.70 | 91.70 |
| ApEn, LogEn, FuzzyEn | 84.30 | 80 | 88.60 |
| LogEn, PermEn, FuzzyEn | 84 | 78.40 | 89.60 |
| PermEn, FuzzyEn, ApEn | 84.25 | 80.40 | 88.10 |
| Four features | |||
| SampEn, ApEn, FuzzyEn, PermEn | 90.25 | 88 | 92.50 |
| SampEn, ApEn, FuzzyEn, LogEn | 89.55 | 87.70 | 91.40 |
| SampEn, FuzzyEn, PermEn, LogEn | 88 | 86.50 | 89.50 |
| SampEn, LogEn, ApEn, PermEn | 89.85 | 88.70 | 91 |
| ApEn, FuzzyEn, PermEn, LogEn | 90.75 | 87.70 | 93.80 |
| Five features | |||
| SampEn, ApEn, FuzzyEn, LogEn, Perm | 92.80 | 91 | 94.60 |