arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1707.09724stat.ML
收藏

带标签噪声的迁移学习

Transfer Learning with Label Noise

Xiyu Yu, Tongliang Liu, Mingming Gong, Kun Zhang, Kayhan Batmanghelich, Dacheng Tao

摘要:

迁移学习的目的是通过从一个相关但不同的源域中借用知识来改善目标域的学习。为了减少源域和目标域之间的分布偏移,最近的方法集中于探索跨域具有相似分布的不变表示。然而,在学习这些不变知识时,现有的方法假设源域中的标签是未污染的,而在现实中,我们经常可以访问带有噪声标签的源数据。在本文中,我们首先展示了标签噪声如何在各种迁移学习场景中对不变表示的学习和标签移位的校正产生不利影响。为了减少不利影响,我们提出了一种新的去噪条件不变分量(DCIC)框架,该框架可证明地确保(1)在源域中具有噪声标签的示例和目标域中未标记的示例中提取不变表示;(2)估计目标域中的标签分布,没有偏见。仿真和真实数据的实验结果验证了该方法的有效性。

英文摘要:

Transfer learning aims to improve learning in target domain by borrowing knowledge from a related but different source domain. To reduce the distribution shift between source and target domains, recent methods have focused on exploring invariant representations that have similar distributions across domains. However, when learning this invariant knowledge, existing methods assume that the labels in source domain are uncontaminated, while in reality, we often have access to source data with noisy labels. In this paper, we first show how label noise adversely affect the learning of invariant representations and the correcting of label shift in various transfer learning scenarios. To reduce the adverse effects, we propose a novel Denoising Conditional Invariant Component (DCIC) framework, which provably ensures (1) extracting invariant representations given examples with noisy labels in source domain and unlabeled examples in target domain; (2) estimating the label distribution in target domain with no bias. Experimental results on both synthetic and real-world data verify the effectiveness of the proposed method.

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。

↑