Link Prediction in Complex Network via Penalizing Noncontribution Relations of Endpoints

Table 2

Prediction accuracy measured by AUC values on the nine benchmark networks. Each data point is an average over 10 independent runs, each of which corresponds to a random 90%–10% division of training set and testing set. All the present results are optimal values. Numbers in brackets stand for the standard deviations.

AUC	CN	PA	AA	RA	LP	BLP	SRW	NRP
USAir	0.937 (0.008)	0.885 (0.017)	0.948 (0.007)	0.954 (0.007)	0.938 (0.007)	0.931 (0.010)	0.953 (0.011)	0.957 (0.011)
Yeast	0.723 (0.005)	0.494 (0.008)	0.724 (0.005)	0.723 (0.005)	0.732 (0.005)	0.733 (0.004)	0.735 (0.005)	0.736 (0.005)
NS	0.940 (0.011)	0.679 (0.012)	0.940 (0.011)	0.940 (0.011)	0.940 (0.011)	0.943 (0.009)	0.943 (0.009)	0.944 (0.009)
Jazz	0.953 (0.005)	0.762 (0.013)	0.961 (0.004)	0.970 (0.004)	0.954 (0.005)	0.951 (0.005)	0.960 (0.004)	0.970 (0.004)
CE	0.914 (0.011)	0.808 (0.021)	0.948 (0.010)	0.954 (0.010)	0.914 (0.011)	0.911 (0.007)	0.953 (0.009)	0.960 (0.009)
Slavko	0.941 (0.009)	0.775 (0.013)	0.945 (0.009)	0.946 (0.010)	0.944 (0.010)	0.943 (0.010)	0.949 (0.009)	0.952 (0.009)
Email	0.844 (0.007)	0.782 (0.007)	0.846 (0.007)	0.846 (0.007)	0.893 (0.006)	0.902 (0.005)	0.903 (0.007)	0.908 (0.006)
Infec	0.939 (0.009)	0.703 (0.013)	0.942 (0.009)	0.943 (0.009)	0.954 (0.012)	0.958 (0.006)	0.964 (0.006)	0.966 (0.006)
ES	0.910 (0.005)	0.820 (0.007)	0.912 (0.006)	0.912 (0.006)	0.936 (0.007)	0.938 (0.005)	0.945 (0.005)	0.946 (0.005)