A Computationally Efficient Method for Hybrid EEG-fNIRS BCI Based on the Pearson Correlation

Table 3

The average classification accuracies of hybrid EEG-fNIRS for eight subjects as calculated through KNN and Tree classifiers.

Feature set fNIRS	Feature set EEG
	,	, SK	, KR	, SK	, KR	SK, KR
,	(73.82, 76.75)	(67.46, 76.45)	(66.85, 75.38)	(67.36, 74.72)	(65.68, 74.85)	(62.08, 73.4)
, SK	(75.28, 76.91)	(69.25, 78.2)	(67.93, 75.5)	(69.43, 76.02)	(67.37, 75.2)	(64.58, 73.7)
, KR	(74.32, 74.53)	(68.98, 74.6)	(67.82, 74.02)	(68.51, 73.62)	(67.31, 73.38)	(63.03, 70.6)
, SK	(73.83, 77.63)	(67.36, 77.4)	(67.5, 77.03)	(67.2, 74.83)	(67.63, 75.76)	(62.55, 73.7)
, KR	(74.2, 76.53)	(69.68, 75)	(68.75, 74.65)	(69.3, 73.6)	(67.6, 73.78)	(63.8, 72.32)
SK, KR	(71.81, 69.88)	(67.03, 69.9)	(65.38, 69.25)	(66.2, 69.2)	(62.16, 69.3)	(60.76, 66.7)