The class imbalance problem: A systematic study

By Nathalie Japkowicz, Shaju Stephen. 2002

3,411 citations recorded by the source corpus.

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

Citations in the corpus

Long-Tailed Class Incremental Learning (2022, methodology): “Learning from long-tailed distributions has been approach by re-sampling [11,15] or re-weighting [21,36,48] head classes and tail classes to learn a balanced classifier.”
Imbalanced Continual Learning with Partitioning Reservoir Sampling (2020, background): “The long-tailed data distribution is both an enduring and pervasive problem in machine learning [40,36], as most real-world datasets are inherently imbalanced [70,63,98,80,65].”
A systematic study of the class imbalance problem in convolutional neural networks (2017, background): “Na¨ıve version, called random majority undersampling, simply removes a random portion of examples from majority classes [17].”
Deep Reconstruction-Classification Networks for Unsupervised Domain Adaptation (2016, background): “Domain adaptation is a large field of research, with related work under several names such as class imbalance [20], covariate shift [21], and sample selection bias [22].”
Predicting Solar Flares Using SDO/HMI Vector Magnetic Data Products and the Random Forest Algorithm (2017, background): “In order to alleviate the class imbalance issue that poses a major challenge in machine learning (e.g., Japkowicz & Stephen 2002), we randomly select 142 unique C-class ARs from their total 552 samples to form a complete data set.”
Training neural network classifiers for medical decision making: The effects of imbalanced datasets on classification performance (2008, background): “A study oriented on isolating the effect of some data properties was presented in Japkowicz (2000) and Japkowicz and Stephen (2002).”
Supervised Machine Learning: A Review of Classification Techniques (2007, background): “Another problem could be that the dataset is imbalanced (Japkowicz & Stephen, 2002).”
Classification with class imbalance problem: A review (2015, background): “Previous work in [37] discovered that as the level of data complexity increases, the class imbalance factor begins to affect the generalization capability of a classifier.”
Minority report in fraud detection: classification of skewed data (2004, methodology): “Skewed data is transformed into partitions with more of the rarer examples and fewer of the common examples using the procedure known as minority oversampling with replacement/replication [24].”
DARE: Data Augmented Relation Extraction with GPT-2 (2020, background): “One of the simplest approaches, random majority undersampling, simply removes a random portion of examples from majority classes so that per class training examples are roughly equal (Japkowicz and Stephen, 2002).”
A study of the behavior of several methods for balancing machine learning training data (2004, background): “On the other hand, Japkowicz and Stephen [13] compare several methods of over and under-sampling on a series of artiﬁcial data sets and conclude that over-sampling is more eﬀective than under-sampling at reducing error rate.”
Balanced Knowledge Distillation for Long-tailed Learning (2021, background): “As opposed to over-sampling, under-sampling [13, 20, 3] randomly discards examples from the majority”