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

Table 2

Results with |L₃| = 153, K = 120, and T = 100.

Method	Feature	HL↓	C↓	OE↓	RL↓	AP↑
RAkEL	LDA	0.0120 ± 0.0003	67.4355 ± 1.5205	0.3044 ± 0.0125	0.2317 ± 0.0057	0.6205 ± 0.0074
	Word vector	0.0114 ± 0.0003	74.6069 ± 1.7410	0.2636 ± 0.0147	0.2527 ± 0.0088	0.6228 ± 0.0068
MLkNN	LDA	0.0113 ± 0.0002	11.3434 ± 0.5319	0.2511 ± 0.0074	0.0347 ± 0.0016	0.6650 ± 0.0065
	Word vector	0.0101 ± 0.0002	11.7522 ± 0.4557	0.2015 ± 0.0101	0.0318 ± 0.0015	0.7289 ± 0.0086
CC	LDA	0.0136 ± 0.0003	71.7498 ± 3.5310	0.4942 ± 0.0107	0.2533 ± 0.0142	0.5108 ± 0.0098
	Word vector	0.0134 ± 0.0002	61.0079 ± 1.8989	0.4427 ± 0.0109	0.2050 ± 0.0075	0.5430 ± 0.0075
BP-MLL	LDA	0.0362 ± 0.0043	10.2577 ± 0.5443	0.2531 ± 0.0087	0.0302 ± 0.0022	0.6522 ± 0.0149
	Word vector	0.0276 ± 0.0011	10.6332 ± 0.4318	0.2417 ± 0.0140	0.0283 ± 0.0009	0.6751 ± 0.0091