Skip to main content

Table 4 Results of Random forest on the missed abortion and diabetes datasets after sampled

From: A hybrid sampling algorithm combining synthetic minority over-sampling technique and edited nearest neighbor for missed abortion diagnosis

 

Algorithm

Original

SM

BSM

ASM

GSM

KSM

CSM

ENN

TL

IHU

RBU

SMTOM

SMENN

Missed abortion

Precision

76.0

76.9

78.0

73.5

84.3

84.0

83.8

81.8

76.0

93.0

63.4

75.5

98.2

Sensitivity

50.0

64.9

72.3

69.6

79.1

79.1

79.1

60.7

50.0

86.6

53.6

68.7

93.9

Specificity

89.2

86.3

83.1

77.1

88.8

88.4

88.0

92.7

89.2

98.2

72.3

81.5

100.0

F-measure

75.9

75.4

77.6

73.3

83.9

83.7

83.5

81.1

75.9

92.4

62.6

75.0

98.1

MCC

43.0

52.5

55.8

46.9

68.2

67.8

67.3

58.1

43.0

85.4

26.4

50.4

95.4

AUC

74.5

81.6

82.2

80.0

86.9

86.0

87.2

82.6

74.5

93.8

57.9

81.2

99.9

Diabetes

Precision

74.8

81.7

82.7

82.1

77.6

81.9

81.5

87.8

79.5

90.4

78.0

84.1

95.1

Sensitivity

61.6

85.8

88.2

87.0

77.6

81.4

82.8

87.7

69.8

87.7

78.0

87.0

97.0

Specificity

82.4

77.2

76.2

76.4

77.6

82.4

80.2

87.9

85.6

92.9

78.0

80.9

92.6

F-measure

75.0

81.5

82.1

81.6

77.6

81.9

81.5

87.8

79.5

90.3

78.0

83.9

95.1

MCC

44.6

63.2

64.9

63.8

55.2

63.8

63.0

75.5

56.2

80.7

56.0

68.0

90.0

AUC

81.9

89.5

89.0

89.2

86.4

90.5

89.7

95.0

85.9

96.2

85.5

91.0

98.9