Table 9.

Classification Results of Emotions in Tablet and Conventional Setups using SVM, RF and LSTM

Item	Metric	NU			Item	Metric	PN
set		SVM	RF	LSTM	set		SVM	RF	LSTM
$T_{\mathit{env}}$	Accuracy	0.983	0.995	0.869	$C_{\mathit{env}}$	Accuracy	0.941	1.000	0.726
	Precision	0.988	0.996	0.869		Precision	0.947	1.000	0.827
	Recall	0.974	0.992	0.894		Recall	0.941	1.000	0.726
	F1-score	0.980	0.994	0.864		F1-score	0.940	1.000	0.695
	Specificity	0.974	0.992	0.894		Specificity	0.941	1.000	0.726
	GSP	0.981	0.994	0.873		GSP	0.942	1.000	0.733
	GSS	0.974	0.992	0.887		GSS	0.938	1.000	0.657
	MCC	0.962	0.988	0.763		MCC	0.888	1.000	0.537
$T_{\mathit{ton}}$	Accuracy	0.667	0.667	0.714	$C_{\mathit{ton}}$	Accuracy	0.385	0.701	0.769
	Precision	0.333	0.333	0.687		Precision	0.206	0.516	0.710
	Recall	0.500	0.500	0.738		Recall	0.308	0.514	0.625
	F1-score	0.400	0.400	0.679		F1-score	0.242	0.438	0.607
	Specificity	0.500	0.500	0.738		Specificity	0.308	0.514	0.625
	GSP	0.408	0.408	0.695		GSP	0.249	0.467	0.635
	GSS	0.000	0.000	0.585		GSS	0.029	0.098	0.402
	MCC	0.000	0.000	0.457		MCC	-0.409	0.082	0.358