normal vs. osteoporosis | osteopenia vs. osteoporosis | normal vs. osteopenia | |||||
---|---|---|---|---|---|---|---|
PN | PP | PN | PP | PN | PP | ||
Support vector machine | |||||||
Confusion matrix | TN | 51 | 2 | 17 | 15 | 48 | 5 |
TP | 1 | 47 | 7 | 41 | 4 | 28 | |
AUC (95%CI) | 0.987(0.964–1.00) | 0.721(0.604–0.839) | 0.962(0.924–1.00) | ||||
Precision | 0.959 | 0.732 | 0.848 | ||||
Recall | 0.979 | 0.854 | 0.875 | ||||
Accuracy | 0.970 | 0.725 | 0.894 | ||||
F1 score | 0.970 | 0.716 | 0.894 | ||||
Random forest | |||||||
Confusion matrix | TN | 52 | 1 | 23 | 9 | 48 | 5 |
TP | 2 | 46 | 8 | 40 | 6 | 26 | |
AUC (95%CI) | 0.994(0.979–1.00) | 0.866(0.779–0.954) | 0.945(0.899–0.992) | ||||
Precision | 0.979 | 0.816 | 0.839 | ||||
Recall | 0.958 | 0.833 | 0.812 | ||||
Accuracy | 0.970 | 0.788 | 0.870 | ||||
F1 score | 0.970 | 0.787 | 0.870 | ||||
K-nearest neighbor | |||||||
Confusion matrix | TN | 48 | 5 | 18 | 14 | 50 | 3 |
TP | 1 | 47 | 3 | 45 | 5 | 27 | |
AUC (95%CI) | 0.970(0.936–1.00) | 0.869(0.783–0.956) | 0.940(0.891–0.989) | ||||
Precision | 0.904 | 0.763 | 0.900 | ||||
Recall | 0.979 | 0.938 | 0.844 | ||||
Accuracy | 0.940 | 0.788 | 0.906 | ||||
F1 score | 0.941 | 0.776 | 0.905 |