Table 11 The optimal classification accuracy due to feature selection (DWT-based FE methods).

From: Selecting EEG channels and features using multi-objective optimization for accurate MCI detection: validation using leave-one-subject-out strategy

FE method

Accuracy (no. of selected features, no. of channel)

RF

SVM

DA

KNN

DWT + STD

91.18 (9, 5)

95.56 (14, 11)

95.42 (8, 8)

91.04 (13, 8)

DWT + ThEn

90.63 (9, 6)

95.83 (12, 8)

95.63 (13, 8)

91.39 (9, 6)

DWT + SuEn

90.28 (14, 10)

95.83 (10, 8)

95.76 (9, 8)

92.92 (16, 9)

DWT + TShEn

90.35 (12, 6)

95.35 (13, 9)

95.42 (14, 10)

91.81 (17, 9)

DWT + KFD

90.14 (18, 12)

95.69 (11, 10)

95.00 (12, 10)

92.50 (17, 10)

DWT + TeEng

90.14 (12, 8)

95.83 (9, 8)

95.83 (12, 9)

93.13 (10, 6)

Back to article page