Skip to main content

Table 2 Classification accuracy for the “Training_Data” set

From: Classification of Parkinson’s disease utilizing multi-edit nearest-neighbor and ensemble learning algorithms with speech samples

 

Leave-one-out method

LOO (%)

Leave-one-subject-out method

LOSO (%)

Accuracy

Sensitivity

Specificity

Accuracy

Sensitivity

Specificity

DNNE (with MENN)

 Mean

67.71

71.71

64

73.37

71

75

 Std

0.037

0.018

0.033

0.042

0.056

0.032

 Best

68.79

72.83

70

85

90

80

RF (with MENN)

 Mean

70.93

73.27

67.59

81.50

92.50

70.50

 Std

0.031

0.0052

0.0063

0.0105

0.0225

0.0211

 Best

76.07

73.65

68.46

88

97

88

SVM (linear)

 Mean

63.65

63.85

63.46

65

65

65

 Std

0

0

0

0

0

0

 Best

63.65

63.85

63.46

65

65

65

SVM (RBF)

 Mean

63.08

73.08

57.08

67.5

80

55

 Std

0

0

0

0

0

0

 Best

63.08

73.08

57.08

67.5

80

55

Method in [1]

 Mean

52.06

54.92

49.22

 Best

85

80

90

Method in [27]

 Mean

 Best

70, KNN (k = 1)

67.5, KNN (k = 3)

72.5, KNN (k = 5)

77.5, KNN (k = 7)

85, SVM (linear)

87.5, SVM (RBF)

80, Naive Bayes

82.5, Discriminant

80, KNN (k = 1)

75, KNN (k = 3)

70, KNN (k = 5)

80, KNN (k = 7)

85, SVM (linear)

90, SVM (RBF)

80, Naive Bayes

80, Discriminant

60, KNN (k = 1)

60, KNN (k = 3)

75, KNN (k = 5)

75, KNN (k = 7)

85, SVM (linear)

85, SVM (RBF)

80, Naive Bayes

85, Discriminant

  1. DNNE (with MENN) and RF (with MENN): reflect the proposed PD_MEdit_EL algorithm; SVM (with linear): SVM with the linear kernel function; SVM (with RBF): SVM with radial basis function kernel; Method in [1]: classification algorithm from the Ref. [1] and Method in [27]: classification algorithm from Ref. [27]