arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1909.03434cs.LGcs.CLcs.SDeess.ASstat.ML
收藏

多标签分类中消除暴露偏差的无序学习

Order-free Learning Alleviating Exposure Bias in Multi-label Classification

National Taiwan University(台湾大学)

Che-Ping Tsai, Hung-Yi Lee

摘要:

多标签分类(MLC)为每个样本分配多个标签。先前的研究表明,MLC可以转换为一个序列预测问题与递归神经网络(RNN)解码器的标签依赖性建模。然而,训练RNN解码器需要预定义的标签顺序,这在MLC规范中并不直接可用。此外,这样训练的RNN倾向于过拟合训练集中的标签组合,并且难以生成看不见的标签序列。在本文中,我们提出了一个新的框架MLC不依赖于一个预定义的标签顺序,从而消除曝光偏差。在三个多标签分类基准数据集上的实验结果表明,该方法的性能大大优于竞争基线。我们还发现,与基线模型相比,所提出的方法生成训练期间未看到的标签组合的概率更高。实验结果表明,该方法具有较好的泛化能力。

英文摘要:

Multi-label classification (MLC) assigns multiple labels to each sample. Prior studies show that MLC can be transformed to a sequence prediction problem with a recurrent neural network (RNN) decoder to model the label dependency. However, training a RNN decoder requires a predefined order of labels, which is not directly available in the MLC specification. Besides, RNN thus trained tends to overfit the label combinations in the training set and have difficulty generating unseen label sequences. In this paper, we propose a new framework for MLC which does not rely on a predefined label order and thus alleviates exposure bias. The experimental results on three multi-label classification benchmark datasets show that our method outperforms competitive baselines by a large margin. We also find the proposed approach has a higher probability of generating label combinations not seen during training than the baseline models. The result shows that the proposed approach has better generalization capability.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑