Research Article
Risk Assessment of Hip Fracture Based on Machine Learning
Table 15
Results of SVM with a Bayesian kernel and posterior probability regions.
| | Mean | St. deviation | Best |
| Sensitivity (8 features) | 56.67% | 2.48% | 55.56% | Specificity (8 features) | 80.00% | 6.88% | 84.62% | Accuracy (8 features) | 66.45% | 2.89% | 67.74% | Sensitivity (12 features) | 71.11% | 3.51% | 77.78% | Specificity (12 features) | 80.77% | 8.31% | 92.31% | Accuracy (12 features) | 75.16% | 4.57% | 83.87% |
|
|