Class Noise vs. Attribute Noise: A Quantitative Study
By Xingquan Zhu, Xindong Wu. Artificial Intelligence Review, 2003
918 citations recorded by the source corpus.
Subfields: Artificial Intelligence, Information Systems
View Paper: https://www.semanticscholar.org/paper/c63a8640e5d426b1c8b0ca2ea45c20c265b3f2ad
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
- Dan Hendrycks
- Xiaogang Wang
- Mantas Mazeika
- Yi Yang
- Taghi M. Khoshgoftaar
- Kevin Gimpel
- Jacob Goldberger
- Jianfeng Gao
In Information Systems
- Albert Orriols-Puig
- Mantas Mazeika
- Chunyan Miao
- Amri Napolitano
- Kevin Gimpel
- Taghi M. Khoshgoftaar
- Michel Verleysen
- Benoît Frénay
Citations in the corpus
- Dealing with noise in defect prediction (2011, neutral): “Zhu and Wu [30] descried a quantitative study of the impact of noisy data on classification accuracies using the UCI machine learning datasets.”
- Training deep neural-networks using a noise adaptation layer (2016, background): “The presence of class label noise inherent to training samples has been reported to deteriorate the performance of even the best classifiers in a broad range of classification problems (Nettleton et al. (2010), Pechenizkiy et al. (2006), Zhu & Wu (2004)).”
- A study of the effect of different types of noise on the precision of supervised learning techniques (2010, methodology): “Recently, several works have studied the effect of noise, including attribute noise and class noise, for specific learners such as C4.5 (Zhu et al. 2003; Zhu and Wu 2004).”
- EEG-Based Emotion Recognition Using Regularized Graph Neural Networks (2019, background): “Prior studies have shown that noisy labels can adversely impact classification accuracy [23].”
- Classification in the Presence of Label Noise: A Survey (2014, background): “In [3], [69] and [301], label noise is introduced using a pairwise scheme.”
- Don't be SCAREd: use SCalable Automatic REpairing with maximal likelihood and bounded changes (2013, neutral): “, [17, 23]), as well as, database constraints (if available) to provide a score Pe(t) ∈ [0.”
- Comparing Boosting and Bagging Techniques With Noisy and Imbalanced Data (2011, background): “the latter one is generally more detrimental to classification performance [ 1 ].”
- Using Trusted Data to Train Deep Networks on Labels Corrupted by Severe Noise (2018, background): “Both natural and malicious label corruptions tend to sharply degrade the performance of classification systems [28].”