The Comparative Experimental Study of Multilabel Classification for Diagnosis Assistant Based on Chinese Obstetric EMRs

Table 3

Results with with |L₄| = 71, K = 120, and D = 100.

Method	Feature	HL↓	C↓	OE↓	RL↓	AP↑
RAkEL	LDA	0.0244 ± 0.0004	26.3255 ± 1.1150	0.2799 ± 0.0123	0.1870 ± 0.0081	0.6575 ± 0.0090
	Word vector	0.0237 ± 0.0004	29.9007 ± 0.8173	0.2391 ± 0.0113	0.2074 ± 0.0071	0.6595 ± 0.0082
MLkNN	LDA	0.0241 ± 0.0003	9.2824 ± 0.2916	0.2498 ± 0.0112	0.0631 ± 0.0026	0.6697 ± 0.0085
	Word vector	0.0214 ± 0.0005	9.0997 ± 0.4973	0.2014 ± 0.0103	0.0547 ± 0.0033	0.7356 ± 0.0088
CC	LDA	0.0288 ± 0.0006	34.4526 ± 1.4447	0.4850 ± 0.0220	0.2729 ± 0.0130	0.5228 ± 0.0125
	Word vector	0.0285 ± 0.0004	30.4830 ± 0.7443	0.4427 ± 0.0109	0.2301 ± 0.0068	0.5509 ± 0.0069
BP-MLL	LDA	0.0458 ± 0.0046	7.4636 ± 0.4216	0.2521 ± 0.0128	0.0462 ± 0.0030	0.7081 ± 0.0098
	Word vector	0.0349 ± 0.0014	7.4289 ± 0.4688	0.2325 ± 0.0131	0.0413 ± 0.0028	0.7413 ± 0.0100