Skip to main content

Table 11 Classification results obtained by employing the top 20 texture, top 20 geometric, and top 20 combined texture and geometric features with six different state-of-the-art classifiers

From: Comparative analysis of proficiencies of various textures and geometric features in breast mass classification using k-nearest neighbor

Classifier

Feature set

Accuracy (%)

Sensitivity (%)

Specificity (%)

k-NN

Texture features (T)

73.9

83.1

62.0

Geometry features (G)

77.4

76.0

78.0

Combined (T&G)

85.2

88.0

82.0

SVM

Texture features (T)

71.3

73.8

68.0

Geometry features (G)

72.2

73.8

70.0

Combined (T&G)

81.7

84.6

78.0

DT

Texture features (T)

68.9

69.2

68.0

Geometry features (G)

71.2

67.6

76.0

Combined (T&G)

71.5

73.8

68.0

NB

Texture features (T)

67.9

64.6

72.0

Geometry features (G)

70.4

72.3

68.0

Combined (T&G)

77.4

75.4

80.0

RF

Texture features (T)

73.9

84.6

60.0

Geometry features (G)

74.1

75.4

72.0

Combined (T&G)

80.6

86.1

74.0

ET

Texture features (T)

74.4

83.0

64.0

Geometry features (G)

76.3

81.5

70.0

Combined (T&G)

81.1

83.0

78.0