Table 4. Parameters tuned for analysis of classifiers.

MLP	P	Activation	Early-stopping	Hidden layersize	Learning rate	Learning rate init	Solver	Toll
	V	ReLU	True	(5,5,5)	Adaptive	0.001	Sgd	0.0001
KNN	P	N_neighbour		Metric		n-jobs		Weights
	V	12		Euclidean		None		Uniform
DT	P	Max_depth		Criteria		Min_sample_leaf
	V	100		Entropy		5
RF	P	Max_depth		Criteria		Min_sample_leaf		n_estimators
	V	100		Entropy		1		200
NN	P	Data division		Training		Performance		Derivative
	V	Random		Levenberg_Marquadt		Mean squared error		Default
FES	P	MFs	Nodes	Linear parameters		Nonlinear parameters		Total Parameters
	V	[3,3]	870	2,132		60		2,192
ARIMA	P	p-value	d-value	q-value		Covariance type		Performance
	V	5	1	0		Opaque		Mean squared error