Skip to main content

Table 5 P-values and classification rates of logistic regression models by detected noise variables in the noise layers for the three-layer structure

From: A biplot correlation range for group-wise metabolite selection in mass spectrometry

δi Level p-value classification rate
BCS FDR1 FDR2 STOC BCS FDR1 FDR2 STOC
0 0.01 0.0001 0.0001 0.6568 0.6600
0.03 0.0002 0.0002 0.0002 0.0002 0.6578 0.6547 0.6526 0.6563
0.05 0.0004 0.0004 0.0004 0.0004 0.6539 0.6508 0.6481 0.6532
0.07 0.0005 0.0005 0.0005 0.6578 0.6537 0.6560
0.10 0.0005 0.0005 0.0005 0.0005 0.6710 0.6669 0.6663 0.6691
0.15 0.0004 0.0004 0.0004 0.0004 0.6769 0.6743 0.6739 0.6767
0.20 0.0001 0.0001 0.0002 0.0001 0.6998 0.6979 0.6958 0.6996
0.03 0.01 0.0097 0.6200
0.03 0.0006 0.0113 0.6450 0.6150
0.05 0.0000 0.0015 0.7200 0.6300
0.07 0.0000 0.0000 0.8200 0.7000
0.10 0.0000 0.0000 0.8250 0.7100
0.15 0.0000 0.0000 0.7900 0.6650
0.20 0.0000 0.0000 0.8150 0.7700
0.05 0.01 0.0535 0.6000
0.03 0.0003 0.6750
0.05 0.0001 0.7150
0.07 0.0000 0.6900
0.10 0.0030 0.6600
0.15 0.0000 0.7550
0.20 0.0000 0.7950