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7 测试数据在Macro F₁上的预测结果

Table 7 Prediction results for test data on Macro F₁

Data set	$ρ$	LEWL	MLML	LRML	CPLST	BR
Emotions	0.3	0.606 $\pm$ 0.057	0.576 $\pm$ 0.051	0.616 $\pm$ 0.036	0.429 $\pm$ 0.069	0.389 $\pm$ 0.021
Emotions	0.7	0.596 $\pm$ 0.038	0.523 $\pm$ 0.061	0.610 $\pm$ 0.036	0.152 $\pm$ 0.033	0.160 $\pm$ 0.041
Yeast	0.3	0.461 $\pm$ 0.004	0.334 $\pm$ 0.011	0.367 $\pm$ 0.008	0.251 $\pm$ 0.009	0.234 $\pm$ 0.009
Yeast	0.7	0.452 $\pm$ 0.006	0.313 $\pm$ 0.008	0.366 $\pm$ 0.009	0.058 $\pm$ 0.009	0.062 $\pm$ 0.004
CAL500	0.3	0.234 $\pm$ 0.012	0.073 $\pm$ 0.007	0.212 $\pm$ 0.011	0.035 $\pm$ 0.005	0.046 $\pm$ 0.003
CAL500	0.7	0.232 $\pm$ 0.008	0.066 $\pm$ 0.008	0.205 $\pm$ 0.010	0.008 $\pm$ 0.002	0.015 $\pm$ 0.002
Medical	0.3	0.358 $\pm$ 0.032	0.311 $\pm$ 0.012	0.325 $\pm$ 0.002	0.310 $\pm$ 0.023	0.098 $\pm$ 0.018
Medical	0.7	0.306 $\pm$ 0.020	0.258 $\pm$ 0.008	0.287 $\pm$ 0.019	0.171 $\pm$ 0.029	0.01 $\pm$ 0.003
Langlog	0.3	0.448 $\pm$ 0.018	0.164 $\pm$ 0.011	0.389 $\pm$ 0.028	0.171 $\pm$ 0.025	0.185 $\pm$ 0.014
Langlog	0.7	0.434 $\pm$ 0.017	0.232 $\pm$ 0.009	0.374 $\pm$ 0.029	0.048 $\pm$ 0.009	0.059 $\pm$ 0.0120.
Enron	0.3	0.166 $\pm$ 0.011	0.065 $\pm$ 0.013	0.120 $\pm$ 0.015	0.063 $\pm$ 0.009	0.042 $\pm$ 0.008
Enron	0.7	0.149 $\pm$ 0.010	0.051 $\pm$ 0.014	0.101 $\pm$ 0.016	0.022 $\pm$ 0.004	0.013 $\pm$ 0.011
Corel5k	0.3	0.025 $\pm$ 0.001	0.018 $\pm$ 0.005	0.020 $\pm$ 0.003	0.010 $\pm$ 0.003	0.009 $\pm$ 0.003
Corel5k	0.7	0.019 $\pm$ 0.001	0.015 $\pm$ 0.006	0.016 $\pm$ 0.003	0.007 $\pm$ 0.004	0.006 $\pm$ 0.002