Table 4 Comparison of different classification algorithms in both men and women to predict healthy individuals, and individuals with osteopenia and OP.

From: Artificial intelligence used to diagnose osteoporosis from risk factors in clinical data and proposing sports protocols

 

Models

AUROC

Accuracy

Precision

Sensitivity

specificity

F-score

Men

DT

0.84

0.81

0.89

0.70

0.86

0.71

RF

0.91

0.85

0.92

0.76

0.89

0.79

KNN

0.86

0.73

0.71

0.70

0.83

0.69

SVM

0.84

0.74

0.83

0.64

0.82

0.65

GB

0.89

0.83

0.90

0.75

0.88

0.77

ET

0.88

0.80

0.88

0.70

0.86

0.73

AB

0.84

0.81

0.89

0.70

0.86

0.71

ANN

0.90

0.73

0.77

0.64

0.81

0.65

Women

DT

0.91

0.91

0.93

0.87

0.93

0.90

RF

0.93

0.92

0.91

0.90

0.95

0.91

KNN

0.87

0.79

0.80

0.76

0.86

0.77

SVM

0.86

0.82

0.83

0.77

0.87

0.79

GB

0.95

0.93

0.93

0.91

0.95

0.92

ET

0.94

0.91

0.93

0.87

0.93

0.90

AB

0.87

0.91

0.93

0.87

0.93

0.90

ANN

0.84

0.75

0.74

0.75

0.82

0.76