Entropy-Based Pattern Learning Based on Singular Spectrum Analysis Components for Assessment of Physiological Signals

Table 7

Performance evaluations for the proposed entropy-based pattern learning to classify the aggressive and normal physical actions from EMG signals. The cross-validation strategy is based on LOOCV. EnM, entropy measures; DimEnM, dimension of entropy measures; DimIF, dimension of input features.

Study	Classifier	EnM	DimEnM	DimIF	Accuracy
Our work	SVM	ApEn	20	8	77.50%
	SVM	SampEn	20	9	80.00%
	SVM	FuzzyEn	20	20	87.50%
	SVM	MSEn	20	7	80.00%
	LDA	ApEn	20	8	72.50%
	LDA	SampEn	20	9	75.00%
	LDA	FuzzyEn	20	20	68.78%
	LDA	MSEn	20	7	73.75%
Pham’s study [14]	LDA	TSME	20	20	61.25%
	LDA	MSEn	20	20	50.00%