Research Article
Risk Assessment of Hip Fracture Based on Machine Learning
Table 17
Results of SVM with a Gaussian kernel and posterior probability regions.
| | Mean | St. deviation | Best |
| | Sensitivity (8 features) | 86.22% | 9.39% | 100.00% | | Specificity (8 features) | 63.38% | 15.35% | 100.00% | | Accuracy (8 features) | 76.65% | 8.97% | 100.00% | | Sensitivity (12 features) | 93.67% | 5.91% | 100.00% | | Specificity (12 features) | 62.92% | 16.56% | 92.31% | | Accuracy (12 features) | 80.77% | 7.57% | 96.77% |
|
|
