一种基于嵌入式的弱标记分类算法

一种基于嵌入式的弱标记分类算法

李亚重,杨有龙,仇海全

Label embedding for weak label classification

Yachong Li,Youlong Yang,Haiquan Qiu

表6 测试数据在平均精度上的预测结果

Table 6 Prediction results for test data on Average Precision

Data set	$ρ$	LEWL	MLML	LRML	CPLST	BR
Emotions	0.3	0.732 $\pm$ 0.026	0.716 $\pm$ 0.01999	0.729 $\pm$ 0.0222	0.662 $\pm$ 0.038	0.645 $\pm$ 0.020
Emotions	0.7	0.708 $\pm$ 0.012	0.687 $\pm$ 0.039	0.682 $\pm$ 0.016	0.558 $\pm$ 0.023	0.568 $\pm$ 0.021
Yeast	0.3	0.708 $\pm$ 0.006	0.654 $\pm$ 0.005	0.441 $\pm$ 0.024	0.587 $\pm$ 0.004	0.573 $\pm$ 0.005
Yeast	0.7	0.698 $\pm$ 0.005	0.648 $\pm$ 0.008	0.44 $\pm$ 0.026	0.448 $\pm$ 0.008	0.450 $\pm$ 0.003
CAL500	0.3	0.365 $\pm$ 0.017	0.31 $\pm$ 0.013	0.202 $\pm$ 0.011	0.248 $\pm$ 0.012	0.248 $\pm$ 0.004
CAL500	0.7	0.369 $\pm$ 0.013	0.307 $\pm$ 0.013	0.197 $\pm$ 0.009	0.170 $\pm$ 0.008	0.175 $\pm$ 0.009
Medical	0.3	0.805 $\pm$ 0.025	0.734 $\pm$ 0.009	0.775 $\pm$ 0.008	0.675 $\pm$ 0.049	0.289 $\pm$ 0.029
Medical	0.7	0.718 $\pm$ 0.039	0.685 $\pm$ 0.018	0.721 $\pm$ 0.038	0.332 $\pm$ 0.030	0.112 $\pm$ 0.011
Langlog	0.3	0.543 $\pm$ 0.019	0.419 $\pm$ 0.016	0.410 $\pm$ 0.017	0.461 $\pm$ 0.048	0.469 $\pm$ 0.019
Langlog	0.7	0.530 $\pm$ 0.016	0.531 $\pm$ 0.019	0.399 $\pm$ 0.017	0.308 $\pm$ 0.029	0.309 $\pm$ 0.011
Enron	0.3	0.536 $\pm$ 0.041	0.395 $\pm$ 0.051	0.348 $\pm$ 0.007	0.331 $\pm$ 0.045	0.296 $\pm$ 0.043
Enron	0.7	0.493 $\pm$ 0.054	0.346 $\pm$ 0.055	0.252 $\pm$ 0.011	0.213 $\pm$ 0.020	0.197 $\pm$ 0.025
Corel5k	0.3	0.038 $\pm$ 0.003	0.027 $\pm$ 0.008	0.027 $\pm$ 0.005	0.030 $\pm$ 0.007	0.027 $\pm$ 0.007
Corel5k	0.7	0.022 $\pm$ 0.002	0.025 $\pm$ 0.007	0.026 $\pm$ 0.004	0.027 $\pm$ 0.007	0.023 $\pm$ 0.007