Table 4 Comparison of classification accuracy and AUC scores across models and Classes.

From: Recognition of anxiety and depression using gait data recorded by the kinect sensor: a machine learning approach with data augmentation

Class

Metric

LDA (%)

Naive Bayes (%)

Multi-class SVM (%)

DNN (%)

Mild

Accuracy (ConfMat)

77.8

69.4

86.7

80.0

AUC

90

81

96

92

Minimal

Accuracy (ConfMat)

42.2

62.8

85.6

51.1

AUC

76

83

94

86

Moderate

Accuracy (ConfMat)

52.2

71.1

84.4

77.8

AUC

80

78

94

88

Severe

Accuracy (ConfMat)

88.3

76.7

89.4

87.8

AUC

92

84

95

92

Back to article page