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 1906.07413cs.LGcs.CVstat.ML
收藏

基于标签分布感知的边际损失的不平衡数据集学习

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss

Stanford University(斯坦福大学);Toyota Research Institute(丰田研究院)

Kaidi Cao, Colin Wei, Adrien Gaidon, Nikos Arechiga, Tengyu Ma

摘要:

当训练数据集存在严重的类不平衡时,深度学习算法可能会表现不佳,但测试标准需要对不太频繁的类进行良好的泛化。我们设计了两种新的方法来提高在这种情况下的性能。首先,我们提出了一个理论上的原则标签分布感知利润(LDAM)损失的动机,最大限度地减少基于利润的推广界。这种损失在训练过程中取代了标准的交叉熵目标,并且可以应用于具有类别不平衡的训练的先前策略,例如重新加权或重新采样。其次,我们提出了一个简单而有效的训练计划,将重新加权推迟到初始阶段之后,允许模型学习初始表示,同时避免与重新加权或重新采样相关的一些复杂性。我们在多个基准视觉任务上测试了我们的方法,包括现实世界的不平衡数据集iNaturalist 2018。我们的实验表明,这些方法中的任何一种单独使用都可以比现有技术有所改进,并且它们的组合可以实现更好的性能提升。

英文摘要:

Deep learning algorithms can fare poorly when the training dataset suffers from heavy class-imbalance but the testing criterion requires good generalization on less frequent classes. We design two novel methods to improve performance in such scenarios. First, we propose a theoretically-principled label-distribution-aware margin (LDAM) loss motivated by minimizing a margin-based generalization bound. This loss replaces the standard cross-entropy objective during training and can be applied with prior strategies for training with class-imbalance such as re-weighting or re-sampling. Second, we propose a simple, yet effective, training schedule that defers re-weighting until after the initial stage, allowing the model to learn an initial representation while avoiding some of the complications associated with re-weighting or re-sampling. We test our methods on several benchmark vision tasks including the real-world imbalanced dataset iNaturalist 2018. Our experiments show that either of these methods alone can already improve over existing techniques and their combination achieves even better performance gains.

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。

↑