Table 5 Model performance comparison of LOOCV results.

From: Upper limb human-exoskeleton system motion state classification based on semg: application of CNN-BiLSTM-attention model

Model

Accuracy (%)

Precision (%)

Recall (%)

F1 score

CNN + LSTM

77.96 ± 10.38

77.91 ± 11.07

77.90 ± 10.52

0.7708 ± 0.1099

DCNN

56.54 ± 24.20

59.67 ± 24.73

57.01 ± 24.50

0.5055 ± 0.2802

LSTM

46.71 ± 11.54

47.45 ± 12.25

46.71 ± 11.34

0.4538 ± 0.1166

SVM

77.65 ± 6.42

78.04 ± 6.50

77.90 ± 6.31

0.7713 ± 0.7070

The proposed model

88.17 ± 5.39

88.76 ± 4.97

88.13 ± 5.47

0.8799 ± 0.5555

Back to article page