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 1912.00646cs.LGstat.ML
收藏

面向不变表示学习的特征发现与分离

Discovery and Separation of Features for Invariant Representation Learning

Information Sciences Institute(信息科学研究所);University of Southern California(南加州大学)

Ayush Jaiswal, Rob Brekelmans, Daniel Moyer, Greg Ver Steeg, Wael AbdAlmageed, Premkumar Natarajan

摘要:

有监督的机器学习模型通常将不相关的干扰因素与预测目标相关联,这损害了泛化。我们提出了一个框架来训练强大的神经网络,通过学习发现和分离数据的预测和滋扰因素,诱导不变性的滋扰。我们提出了一个信息理论制定我们的方法,从中我们得到的培训目标和它与以前的方法的连接。大量数据集的实证结果表明,所提出的框架实现了最先进的性能,并且在训练期间不需要多余的注释。

英文摘要:

Supervised machine learning models often associate irrelevant nuisance factors with the prediction target, which hurts generalization. We propose a framework for training robust neural networks that induces invariance to nuisances through learning to discover and separate predictive and nuisance factors of data. We present an information theoretic formulation of our approach, from which we derive training objectives and its connections with previous methods. Empirical results on a wide array of datasets show that the proposed framework achieves state-of-the-art performance, without requiring nuisance annotations during training.

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。

↑