Skip to main content

Table 8 The proposed cost-sensitive method with other methods in term of PRC

From: Improved cost-sensitive representation of data for solving the imbalanced big data classification problem

Dataset

PRC (%)

S-MVML-LA

GMEB

Method proposed

Iris

79.86 (3)

95.61 (1)

95.14 (2)

Cancer

96.23 (1)

94.83 (3)

95.86 (2)

Ionosphere

67.77 (3)

80.13 (2)

82.31 (1)

1Wine

81.90 (3)

96.49 (2)

98.19 (1)

Pid

49.89 (3)

71.01 (2)

71.42 (1)

Wdbc

86.69 (3)

93.46 (2)

96.20 (1)

Musk

84.55 (2)

80.06 (3)

87.95 (1)

Dermatology-6

70 (3)

95 (1)

88.33 (2)

FuelCons

56.12 (3)

65.67 (2)

68.82 (1)

Movement_libras

47.49 (3)

76.60(2)

79.40 (1)

Sonar

78.90(2)

78.90 (2)

84.67 (1)

SPECTF

50(2.5)

50 (2)

75.93 (1)

Mnist

76.15 (3)

89.74 (2)

91.55 (1)

Colon

84.29 (2)

84.50 (1)

84.29 (2)

Caltch101

46.84 (2)

45.26 (3)

48.25 (1)

DLBCL77

50 (3)

82.70 (2)

95.50(1)

Kddcup-rootkit-imap-vs-back

68.01(3)

98 (1)

95 (2)

Kddcup-buffer-overflow-vs-back

79.27 (3)

100 (1)

100 (1)

Kddcup-guess-passwd-vs-satan

90.65 (3)

99.90 (2)

100 (1)

Kddcup-land-vs-satan

80.93 (3)

100 (1)

100 (1)

Cleveland

20 (3)

28.65 (2)

31.79 (1)

Average Rank

2.69

1.95

1.30

  1. The numbers in bold are the best accuracies achieved on the dataset