Table 3.

Performance of different multilabel learning algorithms.

Group (mean ± std)	ML-kNN	ECC	BSVM	BP-MLL	RANK-SVM	REAL
Average precision	0.759 ± 0.029	0.802 ± 0.016	0.802 ± 0.016	0.540 ± 0.023	0.707 ± 0.022	0.820 ± 0.029
Coverage	0.200 ± 0.023	0.186 ± 0.019	0.174 ± 0.023	0.345 ± 0.039	0.237 ± 0.016	0.160 ± 0.020
Hamming loss	0.167 ± 0.014	0.148 ± 0.016	0.156 ± 0.014	0.304 ± 0.014	0.214 ± 0.014	0.142 ± 0.019
One error	0.375 ± 0.050	0.261 ± 0.024	0.307 ± 0.022	0.755 ± 0.029	0.449 ± 0.034	0.283 ± 0.055
Ranking loss	0.167 ± 0.025	0.190 ± 0.025	0.130 ± 0.017	0.334 ± 0.040	0.206 ± 0.014	0.117 ± 0.018