Table 4 Classification accuracy of off-PD versus HC using KNN classifier with 10 × tenfold CV.

From: Detection of Parkinson’s disease from EEG signals using discrete wavelet transform, different entropy measures, and machine learning techniques

FE methods

Without DWT

With DWT decomposition

With DWT dec. and recon

Energy

90.26 ± 4.53

96.87 ± 1.81

97.00 ± 2.34

LBP

88.96 ± 4.89

97.36 ± 3.15

97.03 ± 2.08

LogEn

91.58 ± 3.07

96.87 ± 2.12

97.94 ± 1.71

ShEn

81.67 ± 3.47

80.37 ± 5.32

80.33 ± 4.61

ThEn

62.20 ± 8.45

99.51 ± 1.11

99.72 ± 0.66

SuEn

86.96 ± 3.98

98.68 ± 1.05

99.66 ± 0.76

NoEn

87.14 ± 3.98

87.13 ± 5.47

89.98 ± 4.07

TShEn

70.77 ± 6.65

98.36 ± 1.35

99.89 ± 0.31

Back to article page