Skip to main content
. 2012 Sep 3;28(18):i487–i494. doi: 10.1093/bioinformatics/bts412

Table 3.

AUC scores on block-wise cross validation experiments

Ratio L1-Log L1-SVM L2-Log L2-SVM KSVM SCCA
1 0.7071±0.0010 0.7061±0.0015 0.7222±0.0009 0.7316±0.0011 0.7325±0.0006 0.7496±0.0042
5 0.7318±0.0004 0.7286±0.0007 0.7368±0.0005 0.7505±0.0005 NA 0.7496±0.0042
10 0.7254±0.0003 0.7339±0.0005 0.7370±0.0004 0.7479±0.0003 NA 0.7496±0.0042
50 0.7243±0.0004 0.7366±0.0004 0.7378±0.0005 0.7479±0.0004 NA 0.7496±0.0042
100 0.7244±0.0005 0.7352±0.0006 0.7361±0.0005 0.7496±0.0003 NA 0.7496±0.0042
ALL 0.7244±0.0004 0.7377±0.0006 0.7371±0.0005 0.7481±0.0004 NA 0.7496±0.0042

The number of negative examples is varied from the same number of positive examples to the number of all possible negative examples. NA means that it was not computationally feasible.