Table 2.

Confusion matrixes for synthetic data sets

A	1		2		3
PCA	P≥0.01	P<0.01	P≥0.01	P<0.01	P≥0.01	P<0.01
i>100	9900	0	9900	0	9900	0
i≤100	0	100	0	100	0	100
Regression	P≥0.01	P<0.01	P≥0.01	P<0.01	P≥0.01	P<0.01
i>100	9900	0	9900	17	9900	48
i≤100	0	100	0	83	0	52
A	4		5		6
PCA	P≥0.01	P<0.01	P≥0.01	P<0.01	P≥0.01	P<0.01
i>100	9900	0	9900	0	9900	0
i≤100	0	100	0	100	0	100
Regression	P≥0.01	P<0.01	P≥0.01	P<0.01	P≥0.01	P<0.01
i>100	9900	64	9900	72	9900	85
i≤100	0	36	0	28	0	15

P-values were adjusted by BH criterion