Table 3.

Results achieved by baseline models and SANE on HDVD dataset.

Model type	Method	ACC. (%)	F1 Score (%)	AUC	AUPR
Similarity-based	SCPMF	66.93 ± 6.31	69.35 ± 4.89	0.8631 ± 0.0456	0.5090 ± 0.0382
	IMCMDA	59.82 ± 6.89	60.41 ± 6.78	0.6423 ± 0.0683	0.1649 ± 0.0463
RL-MF-based	Laplacian	67.36 ± 2.74	65.38 ± 3.13	0.7217 ± 0.0323	0.6743 ± 0.0163
	GF	80.11 ± 2.83	81.10 ± 2.26	0.8541 ± 0.0247	0.7954 ± 0.0201
RL-RW-based	Deepwalk	77.14 ± 3.29	78.23 ± 2.88	0.8066 ± 0.0232	0.7085 ± 0.0166
	Node2Vec	78.57 ± 3.15	79.64 ± 2.69	0.8347 ± 0.0258	0.7600 ± 0.0239
RL-NN-based	LINE	66.26 ± 1.76	65.24 ± 1.56	0.6913 ± 0.0305	0.6950 ± 0.0279
	SDNE	80.55 ± 3.78	80.86 ± 3.42	0.8599 ± 0.0270	0.8091 ± 0.0170
Proposed Model	SANE	81.98$\pm$6.62	83.43 $\pm$ 5.79	0.8961 $\pm$ 0.0335	0.8707 $\pm$ 0.0529