A New Feature Extraction Algorithm Based on Orthogonal Regularized Kernel CCA and Its Application

Table 5

Comparisons of the accuracies of classification for ORCCA and ORKCCA in different feature combinations and regularization parameters.

Feature combinations	ORCCA			ORKCCA
	λ = 10^−5	λ = 10^−3	λ = 1	ζ = 10^−5	ζ = 10^−3	ζ = 1
X^G-X^L	0.9314	0.9300	0.9375	0.9625	0.9681	0.9687
X^G-X^P	0.9230	0.9228	0.9248	0.9511	0.9525	0.9536
X^G-X^Z	0.9180	0.9196	0.9196	0.9482	0.9518	0.9520
X^L-X^P	0.9187	0.9187	0.9190	0.9500	0.9533	0.9545
X^L-X^Z	0.9200	0.9205	0.9235	0.9574	0.9600	0.9615
X^P-X^Z	0.7413	0.7413	0.7525	0.8436	0.8450	0.8450