Table 3.

First experiment: classifiers’ configuration.

Method	Parameters
MLP	$L=0.2045$, $M=0.1909$, $H=59$, $N=5000$, $S=0$
SVM	$C=1000$, $G=\{0.1,0.001,0.01\}$ *, kernel type = RBF
KNN	KNN $=1$, $A=$ Euclidean distance
MNLR	$R=1\times {10}^{-8}$

* Refers to the gamma value for the face, eyes, and mouth evaluation, respectively.