An ECoG-Based Binary Classification of BCI Using Optimized Extreme Learning Machine

Table 3

Objective classification performance of OELM.

Parameters		Evaluation metrics
Activation function	Hidden layer node number	Classification accuracy (%)	Average training time (s)	Average testing time (s)
sin	53	84.62	0.0049	<0.0001
sig	53	83.33	0.0052	0.0012
tan	53	84.62	0.0061	0.0012
asin	53	84.62	0.0049	0.0012
atan	53	84.62	0.0058	0.0006
acos	52	80.77	0.0077	<0.0001
asinh	53	84.62	0.0070	<0.0001
atanh	53	84.62	0.0098	0.0006
tanh	53	84.62	0.0055	0.0012
sinh	53	84.62	0.0077	<0.0001