Link prediction for ex ante influence maximization on temporal networks

Applied Network Science

Table 2 True positive rate (TPR) and relative degree root mean squared error (MSE) for each method and each network

Dataset	Method	TPR	MSE
Synthetic	JC	0.08	0.65
	LogReg	0.04	0.58
	NMF	0.00	1.22
	GNN	0.01	2.24
Reality	JC	0.02	1.26
	LogReg	0.04	1.19
	NMF	0.04	0.99
	GNN	0.04	1.28
Email 4	JC	0.39	0.65
	LogReg	0.22	0.82
	NMF	0.06	1.49
	GNN	0.22	1.38
High School 1	JC	0.00	1.05
	LogReg	0.00	1.02
	NMF	0.08	1.64
	GNN	0.05	1.67
Hospital	JC	0.36	2.38
	LogReg	0.08	1.26
	NMF	0.11	1.42
	GNN	0.27	1.53
Office	JC	0.34	0.86
	LogReg	0.20	0.79
	NMF	0.16	1.19
	GNN	0.15	1.26
CopenB	JC	0.05	2.16
	LogReg	0.02	1.69
	NMF	0.02	2.13
	GNN	0.02	4.56
College	JC	0.14	7.25
	LogReg	0.02	6.09
	NMF	0.00	4.98
	GNN	0.03	19.14

TPR is the proportion of future edges that the method correctly predicted averaged over all time steps (larger is better). MSE is the root mean squared error between the aggregated degrees of the future and predicted networks, divided by the future average degree (smaller is better)