Skip to main content

Table 6 Accuracy, specificity and sensitivity from kNN, SVM and naive Bayes using voice features

From: Telemonitoring Parkinson’s disease using machine learning by combining tremor and voice analysis

ML

Feat.

Male

Female

Both

Algo.

No.

Accu.

Spec.

Sens.

Accu.

Spec.

Sens.

Accu.

Spec.

Sens.

kNN

22

98.3

89.0

100

95.9

93.0

98.0

93.7

93.7

94.6

10

93.0

83.0

96.0

94.6

97.6

99.6

92.1

87.5

93.6

5

93.9

89.0

96.0

94.6

93.0

95.0

93.1

86.4

94.6

SVM

22

92.3

88.0

96.4

95.9

93.3

97.7

89.9

85.0

93.0

10

93.9

88.0

97.0

94.6

93.3

95.5

90.5

89.5

93.6

5

84.3

82.0

90.0

94.6

92.0

96.0

90.5

89.0

94.0

Naive Bayes

22

87.2

84.0

92.0

77.0

73.3

77.3

74.6

72.5

75.0

10

90.4

87.0

94.0

86.5

83.3

90.0

80.1

77.0

87.0

5

94.8

88.0

96.5

87.8

85.0

92.0

81.0

80.5

89.5