Variational Approach for Learning Community Structures

Table 2

Experimental results on Karate dataset.

	NMI ()	VI ()	ACC ()	Q ()	CON ()	TPR ()
Spectral Clustering	0.2297	2.0005	0.7353	0.1127	0.3702	0.7363
Louvain	0.4900	1.5205	0.3235	0.4188	0.2879	0.7333
DeepWalk	0.7198	0.8812	0.9353	0.3582	0.1337	0.9353
node2vec	0.8372	0.8050	0.9706	0.1639	0.4239	0.4549
Stochastic Blockmodel	0.0105	1.1032	0.4412	-0.2084	0.7154	0.4034
Stochastic Blockmodel (D.C)	0.8372	0.8050	0.9706	0.3718	0.1282	0.9412
VGAE∗ + -means	0.6486	0.8189	0.9647	0.3669	0.1295	0.9407
VGAECD∗	1.0000	0.6931	1.0000	0.3582	0.1412	0.9412