Table 9 Classification accuracy of naïve Bayes, k-NN, and SVM over the Australian and Japanese datasets

Australian

Original

2

14

690

67.34%

71.59% (k = 9)

55.73% (γ = 0)

+2D

16

65.02%

72.75% (k = 7)

56.02% (γ = 0)

2D

2

62.12%

71.88% (k = 14)

62.70% (γ = 0)

Japanese

Original

2

15

653

67.18%

69.02% (k = 5)

55.83% (γ = 0)

+2D

17

64.88%

69.63% (k = 5)

55.52% (γ = 0)

2D

2

61.81%

68.40% (k = 9)

62.58% (γ = 0)