Skip to main content

Table 4 Comparisons on the sensitivity, specificity and accuracy of classification among different machine learning approaches. Performances greater than 85% are highlighted in bold

From: A parsimonious approach for screening moderate-to-profound hearing loss in a community-dwelling geriatric population based on a decision tree analysis

 

Sensitivity

Specificity

Accuracy

Better-hearing ear

Training set (Community A)

Decision Tree (DT)

95.35%

86.85%

91.20%

Support Vector Machine (SVM)

100.00%

100.00%

100.00%

Random Forest (RF)

91.32%

94.97%

93.10%

Multilayer Perceptron (MLP)

78.29%

56.98%

67.88%

Test set (Community B)

Decision Tree (DT)

93.50%

90.56%

91.92%

Support Vector Machine (SVM)

100.00%

33.57%

64.29%

Random Forest (RF)

94.31%

92.31%

93.23%

Multilayer Perceptron (MLP)

90.65%

51.05%

69.36%

Worse-hearing ear

Training set (Community A)

Decision Tree (DT)

97.53%

71.54%

89.93%

Support Vector Machine (SVM)

100.00%

100.00%

100.00%

Random Forest (RF)

98.21%

77.51%

92.15%

Multilayer Perceptron (MLP)

99.89%

11.38%

73.99%

Test set (Community B)

Decision Tree (DT)

97.58%

80.69%

91.17%

Support Vector Machine (SVM)

100.00%

34.16%

75.00%

Random Forest (RF)

99.39%

84.65%

93.80%

Multilayer Perceptron (MLP)

100.00%

19.31%

69.36%