From: Logical Analysis of Data (LAD) model for the early diagnosis of acute ischemic stroke
Method | Performance | Training | Validation | Cross validation |
---|---|---|---|---|
C4.5 Decision Trees | Accuracy | 84.5% | 69.3% | 75.9 ± 2.9% |
 | Sensitivity | 90.3% | 76.0% | 80.4 ± 4.8% |
 | Specificity | 78.7% | 62.5% | 71.4 ± 3.5% |
Logistic Regression | Accuracy | 73.8% | 64.8% | 71.1 ± 1.8% |
 | Sensitivity | 64.5% | 52.0% | 59.4 ± 2.7% |
 | Specificity | 83.0% | 77.5% | 82.8 ± 3.3% |
Support Vector Machines | Accuracy | 81.8% | 68.5% | 77.6 ± 3.1% |
 | Sensitivity | 74.2% | 52.0% | 87.50 ± 3.3% |
 | Specificity | 89.4% | 85.0% | 67.80 ± 5.2% |
Multilayer Perceptron | Accuracy | 88.8% | 68.5% | 82.20 ± 2.4% |
 | Sensitivity | 96.8% | 72.0% | 78.00 ± 3.5% |
 | Specificity | 80.9% | 65.0% | 86.40 ± 3.7% |