Link Prediction via Sparse Gaussian Graphical Model

Table 5

Taro dataset results.

Method	Training set ratio
	60%	70%	80%	90%
SGGM,	0.672	0.711	0.736	0.750
SGGM,	0.713	0.747	0.779	0.815
SGGM,	0.721	0.763	0.816	0.857
SGGM,	0.723	0.788	0.825	0.865
CN	0.529	0.545	0.573	0.582
Salton	0.530	0.542	0.571	0.580
Jaccard	0.468	0.476	0.476	0.478
Sørensen	0.523	0.533	0.544	0.552
HPI	0.533	0.545	0.574	0.585
HDI	0.528	0.541	0.568	0.576
LHN	0.532	0.543	0.570	0.579
AA	0.533	0.554	0.589	0.609
RA	0.533	0.554	0.589	0.610
PA	0.559	0.568	0.582	0.607
LNBCN	0.536	0.562	0.616	0.638
LNBAA	0.537	0.561	0.614	0.634
LNBRA	0.536	0.560	0.613	0.634