No Regret Sample Selection with Noisy Labels
No Regret Sample Selection with Noisy Labels
复制标题
DOI：
--
复制
发表时间：
2020-03
期刊：
影响因子：
--
通讯作者：
N. Mitsuo;S. Uchida;D. Suehiro
中科院分区：
文献类型：
--
作者：
N. Mitsuo;S. Uchida;D. Suehiro
深度神经网络(DNN)由于存在过拟合的风险，容易受到噪声的影响。为了避免这种风险，在本文中，我们提出了一种新的学习噪声样本的样本选择框架。其核心思想是采用一种“后悔”最小化的方法。所提出的样本选择方法自适应地选择带噪声标签的训练样本的子集，以最大限度地减少选择噪声样本的遗憾。该算法运行效率高，具有一定的理论支持。此外，与典型方法不同的是，该算法不需要依赖于DNN训练设置的任何辅助信息或学习信息。实验结果表明，该方法提高了含有噪声标签数据的黑盒DNN的性能。
The Deep Neural Network (DNN) suffers from noisy labeled data because of the risk of overfitting. To avoid the risk, in this paper, we propose a novel sample selection framework for learning noisy samples. The core idea is to employ a "regret" minimization approach. The proposed sample selection method adaptively selects a subset of noisy labeled training samples to minimize the regret of selecting noise samples. The algorithm works efficiently and performs with theoretical support. Moreover, unlike the typical approaches, the algorithm does not require any side information or learning information that depends on the training settings of the DNN. The experimental results demonstrate that the proposed method improves the performance of a black-box DNN with noisy labeled data.