Table 3 Comparison of MaeanAcc and MeanFeat for DESFO-K-NN & the basic K-NN.

From: An improved Differential evolution with Sailfish optimizer (DESFO) for handling feature selection problem

Benchmarks

Accuracy

Features

Basic K-NN

DESFO-

K-NN(Mean)

Increasing

(%)

Basic K-NN

DESFO-

K- NN(Mean)

Decreasing

(%)

PenglungEW

0.5333

0.6533

22.50

325

108.2

66.71

IonosphereEW

0.8451

0.9324

10.33

34.0

9.700

71.47

SonarEW

0.8571

0.9857

15.00

60.0

23.60

60.67

WaveformEW

0.804

0.8480

5.470

40.0

23.20

42.00

KrVsKpEW

0.9656

0.9822

1.720

36.0

20.10

44.17

BreastEW

0.9211

0.9649

4.760

30.0

6.300

79.00

Lymphography

0.7000

0.8400

20.00

18.0

8.500

52.78

Vote

0.9333

1.0000

7.150

16.0

4.100

74.38

Zoo

0.9048

1.0000

10.52

16.0

5.200

67.50

Exactly2

0.7400

0.7935

7.230

13.0

6.300

51.54

M-of-n

0.8800

1.0000

13.64

13.0

6.100

53.08

WineEW

0.5833

1.0000

71.44

13.0

4.000

69.23

BreastCancer

0.6214

0.9857

58.63

9.00

6.000

33.33

Tic tac toe

0.8441

0.8542

1.200

9.00

9.000

0.00

W

Score T

L

0

14

*

0

13

*

 

0

0

1

1

 

14

0

13

0

  1. Superior values are in [bold].
Back to article page