BMC Bioinformatics

Table 3 The overall experimental results on baselines, MuFRF, and the ablation study of MuFRF

From: Multi-view feature representation and fusion for drug-drug interactions prediction

Method	Binary-class				Multi-class
Method	Accuracy	Precision	Recall	F1	Mean-Accuracy	Marco-Precision	Macro-Recall	Macro-F1
DeepWalk	0.8130	0.7970	0.8390	0.8170	0.8000	0.8220	0.7101	0.7469
LINE	0.7810	0.7710	0.8000	0.7850	0.7506	0.6870	0.5451	0.5804
DeepDDI	0.9166	0.9121	0.9241	0.9167	0.8768	0.7986	0.7593	0.7662
KGDDI	0.8926	0.8936	0.8925	0.8925	0.8923	0.7945	0.7667	0.7666
KGNN	0.9034	0.9058	0.8999	0.9029	0.9127	0.8583	0.8170	0.8291
MUFFIN	0.9913	0.9912	0.9913	0.9912	0.9648	0.9568	0.9482	0.9495
RANEDDI	–	–	–	–	0.9707	0.9260	0.9096	0.9161
GRPMF	–	–	–	–	0.9617	0.9627	0.9617	0.9622
MuFRF	0.9945	0.9945	0.9945	0.9945	0.9721	0.9765	0.9676	0.9699
MuFRF_ST	0.9509	0.9510	0.9509	0.9509	0.9524	0.9344	0.9254	0.9248
MuFRF_KG	0.9200	0.9200	0.9199	0.9199	0.9355	0.8971	0.8641	0.8724
MuFRF-c0	0.9840	0.9840	0.9840	0.9840	0.9705	0.9669	0.9570	0.9594
MuFRF-c1	0.9823	0.9823	0.9822	0.9822	0.9718	0.9754	0.9636	0.9667
MuFRF-add	0.9823	0.9823	0.9823	0.9823	0.9730	0.9684	0.9616	0.9625
MuFRF-p	0.9814	0.9815	0.9814	0.9814	0.9703	0.9662	0.9556	0.9581
MuFRF-attn	0.9821	0.9821	0.9821	0.9821	0.9700	0.9652	0.9574	0.9585

The bold values represent the performance of our proposed model MuFRF outperforms all baselines on each metric for binary and multi-class classification tasks

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com