Table 7. Logistic Regression (LR): Hyperparameters’ domain and the corresponding tuned values at the data sets under consideration.

The C, N_i, l₁ ratio, in respect, represents the inverse of regularization strength, the number of iterations, and the Elastic-Net mixing parameter.

method / data set	parameters
	intercept	C	N i	l1 ratio	penalty options
LR	{False, True}	[1e-1, 4]	{100, 101, …, 100000 }	[1e-1, 9e-1]	{none, l1, l2, Elastic Net}
LR at Demo	True	0.4078	29307	0.1445	l2
LR at Fixation	True	0.1302	27169	0.2779	none
LR Demo-Fixation	True	0.1035	85814	0.8695	l1
LR at IA	True	3.123	83030	0.663	l2
LR at Demo-IA	False	0.317	90860	0.611	none