Myoelectric Pattern Recognition Performance Enhancement Using Nonlinear Features

Table 2

The p values between the proposed feature extraction method and each of the existing feature extraction methods.

Parameter	Classifier	FS1	FS2	FS3	FS4
Accuracy	LDA	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	SVM	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	KNN	p < 0.001	p < 0.001	p < 0.001	p < 0.001
Sensitivity	LDA	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	SVM	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	KNN	p < 0.001	p < 0.001	p < 0.001	p < 0.001
Specificity	LDA	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	SVM	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	KNN	p < 0.001	p < 0.001	p < 0.001	p < 0.001
Precision	LDA	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	SVM	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	KNN	p < 0.001	p < 0.001	p < 0.001	p < 0.001
F1 score	LDA	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	SVM	p < 0.001	p < 0.001	p < 0.001	p < 0.001
	KNN	p < 0.001	p < 0.001	p < 0.001	p < 0.001