Recycling weak labels for multiclass classification

Recycling weak labels for multiclass classification
复制标题

回收弱标签进行多类分类

DOI:
10.1016/j.neucom.2020.03.002
复制
发表时间:
2020
期刊:
影响因子:
6
通讯作者:
Perello-Nieto M
Perello-Nieto M
中科院分区:
计算机科学2区
文献类型:
--
作者:
Perello-Nieto M

文献摘要

参考文献

被引文献

相似文献

本文探讨了对多类分类数据集有效地组合不同质量的标注的机制,因为我们认为,相对于真正的标签,更容易获得大的弱标签集合。由于标签来自不同的来源,它们的标注可能具有不同程度的可靠性(例如,噪声标签、标签超集、补充标签或领域专家执行的标注),我们必须确保添加潜在不准确的标签不会降低仅使用真实标签时的性能。为此,我们认为每组标注都是弱监督的,并提出了寻找这些集合的最优组合的问题。我们提出了一种基于期望最大化的高效算法,并在各种弱标签场景下展示了其在合成和真实世界分类任务中的性能。
This paper explores the mechanisms to efficiently combine annotations of different quality for multiclass classification datasets, as we argue that it is easier to obtain large collections ofweaklabels as opposed to true labels. Since labels come from different sources, their annotations may have different degrees of reliability (e.g., noisy labels, supersets of labels, complementary labels or annotations performed by domain experts), and we must make sure that the addition of potentially inaccurate labels does not degrade the performance achieved when using only true labels. For this reason, we consider each group of annotations as being weakly supervised and pose the problem as finding the optimal combination of such collections. We propose an efficient algorithm based on expectation-maximization and show its performance in both synthetic and real-world classification tasks in a variety of weak label scenarios.
DOI: 10.5555/1953048.2021040
发表时间: 2011-01
期刊: J. Mach. Learn. Res.
影响因子: --
作者:
Shohei Shimizu;Takanori Inazumi;Yasuhiro Sogawa;Aapo Hyvärinen;Y. Kawahara;T. Washio;P. Hoyer;K. Bollen
通讯作者: Shohei Shimizu;Takanori Inazumi;Yasuhiro Sogawa;Aapo Hyvärinen;Y. Kawahara;T. Washio;P. Hoyer;K. Bollen
DOI: --
发表时间: 2017
期刊: International Symposium on Intelligent Data Analysis
影响因子: --
作者:
Miquel Perello;Raúl Santos;Jesús Cid
通讯作者: Jesús Cid
向不精确的老师学习:概率和证据方法
DOI: --
发表时间: 2015
期刊:
影响因子: --
作者:
C. Ambroise;T. Denœux;G. Govaert;P. Smets
通讯作者: P. Smets