Skip to main content
. 2015 Sep 9;5:13867. doi: 10.1038/srep13867

Table 1. Performance of RLS-KP by 10-fold cross validation with S d  = S chem .

Fold Id Original algorithm Algorithm with eigenvalue transformation applied p
Training set
Test set
Training set
Test set
α
AUC AUPR AUC AUPR AUC AUPR AUC AUPR
1 93.0 42.9 92.8 51.2 93.5 56.7 94.3 68.9 0.5 0
2 92.5 42.2 92.7 52.9 93.7 56.8 93.7 66.5 0.6 0
3 92.8 42.9 92.7 50.0 93.5 57.4 93.3 65.9 0.5 0
4 92.3 43.3 93.3 44.9 93.6 58.0 93.9 63.7 0.5 0
5 92.6 43.0 92.8 49.5 93.4 57.5 94.8 68.6 0.5 0
6 92.4 43.0 93.0 44.0 93.4 57.6 94.0 60.1 0.6 0
7 92.7 44.0 92.9 42.4 93.6 58.6 94.1 62.9 0.5 0
8 92.4 43.6 91.9 44.5 93.5 58.9 93.6 59.4 0.5 0
9 92.5 42.9 93.8 50.2 93.7 57.3 94.5 65.3 0.5 0
10 92.2 43.1 94.3 51.9 93.4 57.0 94.4 68.1 0.5 0
average 92.5 43.1 93.0 48.1 93.5 57.6 94.1 65.0   0

The AUC scores and AUPR scores are normalized to 100. The p indicates p-value of bootstrapping.