Predicting drug side effects by multi-label learning and ensemble learning

BMC Bioinformatics

Table 7 Performances of the proposed ensemble methods and state-of-the-art methods

Dataset	Method	AUPR	Hamming loss	Ranking loss	One error	Coverage	Average precision
Pauwels’s dataset	Pauwels’s method	0.3883	0.0577	0.0827	0.1779	832.7827	0.4616
Pauwels’s dataset	ensemble method	0.4286	0.0454	0.0737	0.1689	790.6261	0.4925
Mizutani’s dataset	Mizutani’s method	0.4107	0.0557	0.0888	0.1854	862.9757	0.4795
Mizutani’s dataset	ensemble method	0.4504	0.0500	0.0761	0.1657	809.6672	0.5012
Liu’s dataset	Liu's method	0.2514	0.0721	0.0927	0.9291	837.4579	0.2610
Liu’s dataset	ensemble method	0.4802	0.0524	0.0703	0.1202	795.9435	0.5134

ISSN: 1471-2105