Table 5.

Parameter settings of five network representation learning methods

ID	Method	Parameters
1	Node2VEC	walk-length = 80, number-walks = 10, p = 8.0, q = 1.0, dimensions = 64
2	DeepWalk	walk-length = 80, number-walks = 10, dimensions = 64
3	HOPE	dimensions = 64
4	LINE	epoch = 5, order = 3, clf-ratio = 0.5, dimensions = 64
5	SDNE	alpha = 1e-6, beta = 5, nu1 = 1e-5, nu2 = 1e-4, batch_size = 200, epoch = 5, learning_rate = 0.01, dimensions = 64