Table 2 Cross-validation results for classifiers based on the ANN model.

From: Analysis of the retraining strategies for multi-label text message classification in call/contact center systems

Model name

Reference model

STRATEGY 1

STRATEGY 2

Accuracy

SD

Emotica

SD

Accuracy

SD

Emotica

SD

Accuracy

SD

Emotica

SD

[%]

[%]

[%]

[%]

[%]

[%]

[%]

[%]

[%]

[%]

[%]

[%]

Training data

 ANN-PCA

90.55

0.29

70.39

0.96

95.70

0.46

84.08

1.35

97.54

0.07

89.90

0.36

 ANN-LSA

91.76

0.32

74.47

1.14

97.02

0.22

88.54

0.80

97.33

0.14

89.66

0.47

 ANN-ICA

92.22

0.40

75.34

0.80

96.05

0.39

85.24

1.20

97.06

0.09

88.01

0.49

Testing data

 ANN-PCA

90.55

0.58

70.39

1.92

91.26

0.71

72.52

2.03

92.43

0.49

76.70

1.56

 ANN-LSA

91.76

0.64

74.47

2.29

93.02

0.51

77.38

1.28

92.62

0.28

76.41

0.45

 ANN-ICA

92.22

0.80

75.34

1.61

92.93

0.56

76.90

1.46

92.78

0.60

76.99

1.57

  1. Significant values are in bold.
  2. SD Standard deviation.
Back to article page