Skip to main content

Table 7 The proposed cost-sensitive method with other methods in term of RCL

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

Dataset

RCL (%)

S-MVML-LA

GMEB

Proposed cost sensitive

Iris

84.40 (3)

95.14 (2)

96.52 (1)

Cancer

96.28 (1)

94.90 (3)

96.02 (2)

Ionosphere

68.14 (3)

82.37 (2)

83.97 (1)

Wine

90.04 (3)

96.49 (2)

98.55 (1)

Pid

32.52 (3)

71.01 (2)

76.55 (1)

Wdbc

92.54 (3)

92.55 (2)

96.00 (1)

Musk

85.75 (2)

81.19 (3)

87.69 (1)

Dermatology-6

98.15 (1)

98.15 (1)

89.85 (3)

FuelCons

54.29 (3)

77.67 (2)

80.88 (1)

Movement_libras

44.70 (3)

79.72(1)

79.04 (2)

Sonar

82.35 (3)

86.99 (2)

85.41 (1)

SPECTF

39.69 (2)

39.69 (2)

67.38 (1)

Mnist

79.28 (3)

91.00 (2)

91.84 (1)

Colon

82.63 (2)

71.71 (3)

83.21 (1)

Caltch101

56.67 (1)

50.14 (3)

52.69 (2)

DLBCL77

37.66 (3)

78.94 (2)

98.32(1)

Kddcup-rootkit-imap-vs-back

71.92 (3)

99.97 (1)

99.97 (1)

Kddcup-buffer-overflow-vs-back

82.03 (3)

100 (1)

100 (1)

Kddcup-guess-passwd-vs-satan

83.68 (3)

96.83 (2)

100 (1)

Kddcup-land-vs-satan

80.93 (3)

100 (1)

100 (1)

Cleveland

10.82 (3)

22.34 (2)

30.63 (1)

Average rank

2.61

2.04

1.30

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