Research Article
Risk Assessment of Hip Fracture Based on Machine Learning
Table 13
Results of SVM with a sigmoid kernel.
| Sigmoid | Mean | St. deviation | Best |
| Sensitivity (8 features) | 56.33% | 12.61% | 77.78% | Specificity (8 features) | 60.62% | 14.35% | 84.61% | Accuracy (8 features) | 58.13% | 9.17% | 80.65% | Sensitivity (12 features) | 56.50% | 12.41% | 66.67% | Specificity (12 features) | 60.77% | 14.36% | 100.00% | Accuracy (12 features) | 58.29% | 9.16% | 80.65% |
|
|