arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2401.13360cs.LG
收藏

理解和缓解带噪标签学习中样本选择偏差

Understanding and Mitigating the Bias in Sample Selection for Learning with Noisy Labels

College of Computing and Data Science, Nanyang Technological University(南洋理工大学计算与数据科学学院);Singapore University of Technology and Design(新加坡科技设计大学);School of Software, Zhejiang University(浙江大学软件学院)

Qi Wei, Lei Feng, Haobo Wang, Bo An

AI总结:

本文针对带噪标签学习中样本选择存在的数据偏差和训练偏差,提出噪声容忍专家模型ITEM,通过多专家鲁棒网络和混合采样策略实现去偏学习,实验验证其有效性。

摘要:

带噪标签学习旨在给定标签损坏的训练集时保证模型泛化能力。样本选择策略通过选择一个标签可靠的子集进行模型训练,取得了有前景的性能。本文通过实验揭示,现有样本选择方法同时受到数据偏差和训练偏差的影响,分别表现为所选集合不平衡和实际中的累积误差。然而,以往研究仅处理了训练偏差。为解决这一局限,我们提出了一种噪声容忍专家模型(ITEM),用于样本选择中的去偏学习。具体而言,为缓解训练偏差,我们设计了一种集成多个专家的鲁棒网络架构。与主流的双分支网络相比,我们的网络通过集成这些专家,在训练参数更少的情况下表现出更好的选择和预测性能。同时,为缓解数据偏差,我们提出了一种基于两个加权数据采样器的混合采样策略。通过在两个类别判别性小批量的混合数据上训练,模型缓解了不平衡训练集的影响,同时避免了采样策略容易导致的稀疏表示。大量实验和分析证明了ITEM的有效性。我们的代码可在该网址获取:https://github.com/1998v7/ITEM。

英文摘要:

Learning with noisy labels aims to ensure model generalization given a label-corrupted training set. The sample selection strategy achieves promising performance by selecting a label-reliable subset for model training. In this paper, we empirically reveal that existing sample selection methods suffer from both data and training bias that are represented as imbalanced selected sets and accumulation errors in practice, respectively. However, only the training bias was handled in previous studies. To address this limitation, we propose a noIse-Tolerant Expert Model (ITEM) for debiased learning in sample selection. Specifically, to mitigate the training bias, we design a robust network architecture that integrates with multiple experts. Compared with the prevailing double-branch network, our network exhibits better performance of selection and prediction by ensembling these experts while training with fewer parameters. Meanwhile, to mitigate the data bias, we propose a mixed sampling strategy based on two weight-based data samplers. By training on the mixture of two class-discriminative mini-batches, the model mitigates the effect of the imbalanced training set while avoiding sparse representations that are easily caused by sampling strategies. Extensive experiments and analyses demonstrate the effectiveness of ITEM. Our code is available at this url \href{https://github.com/1998v7/ITEM}{ITEM}.

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。

↑