Double-Criteria Active Learning for Multiclass Brain-Computer Interfaces

Table 2

Details of the optimal parameter settings for the different datasets using four methods.

Dataset	D-AL-ELM		AL-ELM	ELM-Entropy	PL-ELM
	Liver	110	0.1	110	110	110
Diabetes	110	0.3	110	110	110
Wdbc	200	0.1	200	200	200
Twonorm	120	0.1	120	120	120
Hayes-Roth	100	0.3	100	100	100
Iris	170	0.9	170	170	170
Wine	60	0.7	60	60	60
Segment	200	0.6	200	200	200
Letter	700	0.4	700	700	700