Classification in the Presence of Label Noise: A Survey

By Benoît Frénay, Michel Verleysen. 2014

1,820 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Industrial and Manufacturing Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computational Theory and Mathematics

In Industrial and Manufacturing Engineering

Citations in the corpus

Support vector machines under adversarial label contamination (2015, background): “While stochastic label noise has been widely studied in the machine learning literature, to account for different kinds of potential labeling errors in the training data [31, 32], only a few works have considered adversarial, worst-case label noise, either from a more theoretical [33] or practical p”
On Robustness of Neural Architecture Search Under Label Noise (2020, background): “Label noise, which corrupts the labels of training instances, has been widely investigated due to its unavoidability in real-world situations and harmfulness to classifier learning algorithms (Frénay and Verleysen, 2013).”
A Survey on Data Collection for Machine Learning: A Big Data - AI Integration Perspective (2018, background): “Hence, there has been extensive research (see the survey [181]) on how to make the model training still use noisy labels by becoming more robust.”
Learning from Noisy Large-Scale Datasets with Minimal Supervision (2017, background): “For a comprehensive overview of label noise taxonomy and noise robust algorithms we refer to [11].”
Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels (2018, background): “Classification with noisy labels is a widely studied topic [8].”
Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe Noise (2018, background): “Accordingly, modeling, correcting, and learning with noisy labels has been well studied [14, 1, 2].”
Robot Learning in Homes: Improving Generalization and Reducing Dataset Bias (2018, background): “However, as the work in Frénay and Verleysen [37] points out, the noise can be either independent of the environment or statistically dependent on the environment.”
Robust support vector machines based on the rescaled hinge loss function (2017, methodology): “In order to study the robustness of the different classifiers, we randomly introduce label noise into the training and validation datasets [4,5].”
A Survey of Visual Transformers (2021, neutral): “Furthermore, it is worthy to study a problem whether co-training [99] can be used as a new way to integrate inductive bias for Transformer in semi-supervision.”
Operational High Resolution Land Cover Map Production at the Country Scale Using Satellite Image Time Series (2017, background): “Random forests (RF) [43] are known to be robust to label noise [44,45] and have been chosen in our approach.”
Robust Loss Functions under Label Noise for Deep Neural Networks (2017, neutral): “A detailed survey is available in [9].”
Federated learning for medical image analysis: A survey (2023, background): “Learning from noisy labels (Song et al., 2022; Frénay and Verleysen, 2013) is the corresponding technique.”
Hyperparameter Selection under Localized Label Noise via Corrupt Validation (2017, background): “Frenay and Verleysen [4] present three main approaches to handling label noise: (1) robust methods that are naturally robust to label noise but do not explicitly model label noise, (2) data cleansing methods that attempt to identify and remove corrupt labels before classification, and (3) noise-tole”
How Complex Is Your Classification Problem? (2018, methodology): “The data complexity measures have also been used to guide data pre-processing tasks, such as Feature Selection (FS) [Liu et al., 2010], noise identiﬁcation [Frenay and Verleysen, 2014] and dealing with data imbalance [He and Garcia, 2009].”