Table 1.

Search spaces of SVM and RF.

Parameters	SVM 1	Parameters	RF
“kernel”	Linear, RBF, sigmoid, poly	“criterion”	Gini, entropy
“C”	2x, x ∈ [−1, 15]	“max_depth”	[5, 150]
“gamma”	2x, x ∈ [−14, 2]	“min_samples_split”	[2, 30]
“degree”	[1, 5]	“min_samples_leaf”	[5]
\	\	“max_leaf_nodes”	[100]
\	\	“ccp_alpha”	[0.001]
\	\	“n_estimators”	10x, x ∈ [1, 3]

¹ When “kernel” is linear, there are no “gamma” and “degree” parameters to be set. When the only ”kernel” specifies as poly, the “degree” parameter makes sense.