Skip to main content

Table 10 Average runtime of the proposed method compared with other methods (in seconds)

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

Dataset

S-MVML-LA

GMEB

IGIS

Proposed

Iris

0.8

0.6

3.2

1.2

Cancer

1.8

0.4

1.3

4.2

Ionosphere

0.9

0.8

4.0

6.4

Wine

0.7

0.5

2.2

2.7

Pid

1.1

0.3

1.6

3.7

Wdbc

1.6

0.6

3.3

8.2

Musk

3.1

1.8

37.2

288.7

Dermatology-6

0.8

0.5

3.3

6.3

FuelCons

3.4

3.2

15.5

84.9

Movement_libras

6.2

14.5

319.0

219.6

Sonar

0.8

0.4

7.5

10.6

SPECTF

0.8

0.4

4.1

10.9

Mnist

1751.7

1767.2

168,000

1461.1

Colon

19.3

3.0

2310.8

204.6

Caltch101

2247.0

9794.4

840,000

94.0

DLBCL77

28.4

23.4

1958.9

20.3

Kddcup-rootkit-imap-vs-back

11.0

1.3

14.0

85.3

Kddcup-buffer-overflow-vs-back

2.7

0.8

7.9

87.7

Kddcup-guess-passwd-vs-satan

2.8

0.7

7.0

59.9

Kddcup-land-vs-satan

4.5

0.7

6.9

48.3

Cleveland

1.7

0.9

5.4

7.2

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