Model-based learning of information diffusion in social media networks

Applied Network Science

Table 2 Performance of large-scale synthetic data

Network type	Model	Nodes	Edges	Cascades	Training accuracy	Testing accuracy
Erdos-Renyi	LTNN	100	270	200	70.5%	66.7%
	RWNN	100	270	200	73.9%	67.5%
	LTNN	1000	24980	200	74.6%	67.6%
	RWNN	1000	24980	200	68.5%	68.0%
Barabasi-Albert	LTNN	100	291	200	69.6%	67.7%
	RWNN	100	291	200	75.1%	67.0%
	LTNN	1000	29100	200	74.5%	68.2%
	RWNN	1000	29100	200	68.7%	68.4%
Watts-Strogatz	LTNN	100	300	200	69.9%	66.9%
	RWNN	100	300	200	74.1%	67.0%
	LTNN	1000	3000	200	69.8%	67.5%
	RWNN	1000	3000	200	73.8%	67.9%