Table 2.

Confusion Matrix and Evaluation Index.

Confusion Matrix		Predicted Label
		Normal	Depression
True Label	Normal	True Positive (TP)	False Negative (FN)
	Depression	False Positive (FP)	True Negative (TN)
Evaluation Index	(1) $\mathrm{\text{Accuracy}}=\frac{TP+TN}{TP+TN+FP+FN}$.
	(2) $\mathrm{\text{Precision}}=\frac{TP}{TP+FP}$.
	(3) $\mathrm{\text{Recall}}=\frac{TP}{TP+FN}$.
	(4) $\mathrm{\text{F1-Score}}=\frac{2\times Precision\times Recall}{Precision+Recall}$.
	(5) $\mathrm{\text{Kappa}}=\frac{P_a-P_e}{1-P_e}$, $P_a=\frac{TP+TN}{TP+TN+FP+FN}$,
	$P_e=\frac{(TP+FP)(TP+FN)+(FN+TN)(FP+TN)}{{(TP+TN+FP+FN)}^2}$.