Skip to main content

Table 1 Test results for accuracy and time required

From: Determining threshold value on information gain feature selection to increase speed and prediction accuracy of random forest

Dataset

CBFS Best First

Threshold 0,05

Treshold based on Standard Deviation

CBFS Best First (FFT)

Threshold 0.05 (FFT)

Treshold based on Standard Deviation (FFT)

Treshold based on Standard Deviation – FFT Fitur with Imaginary

Treshold based on Standard Deviation – FFT Fitur without Imaginary

Accuracy

Time

Accuracy

Time

Accuracy

Time

Accuracy

Time

Accuracy

Time

Accuracy

Time

Accuracy

Time

Accuracy

Time

EEG Eye

77.01%

4.17

63.16%

5.31

90.15%

4.91

77.17%

7.31

72.06%

4.96

90.14%

9.41

62.32%

0.82

62.32%

0.82

Cancer

95.79%

0.08

96.63%

0.09

94.39%

0.07

95.68%

0.08

96.68%

0.10

94.41%

0.08

92.55%

0.05

87.94%

0.02

Contraceptive Method

48.74%

0.25

48.74%

0.25

51.64%

0.25

51.74%

0.35

50.90%

0.21

51.74%

0.27

45.41%

0.08

48.06%

0.05

Dermatology

94.92%

0.07

97.05%

0.05

97.43%

0.04

97.70%

0.07

97.35%

0.07

97.40%

0.07

54.29%

0.02

50.27%

0.01

Divorce

96.53%

0.03

97.65%

0.03

97.65%

0.02

96.65%

0.02

97.71%

0.02

97.65%

0.03

96.47%

0.01

97.00%

0.01

Electrical Grid

100.00%

0.91

100.00%

0.52

100.00%

0.17

85.64%

5.06

76.73%

2.57

80.85%

3.44

100%

1.22

100%

1.22

CNAE-9

81.18%

0.27

87.56%

0.51

88.05%

0.58

81.16%

0.27

90.69%

0.76

90.49%

0.91

21.84%

0.40

76.66%

0.31

Urban Land Cover

87.68%

0.06

85.71%

0.09

84.76%

0.07

87.62%

0.04

85.89%

0.07

84.64%

0.05

29.17%

0.04

56.01%

0.05

Epilepsy

69.48%

15.72

69.60%

26.91

69.60%

26.91

69.51%

21.52

69.73%

27.93

69.73%

27.93

32.46%

0.90

32.46%

0.86

SCADI

84.14%

0.02

83.43%

0.03

83.86%

0.01

85.86%

0.02

85.00%

0.02

85.00%

0.02

41.43%

0.01

61.43%

0.01