A Review on Ensembles for the Class Imbalance Problem: Bagging-, Boosting-, and Hybrid-Based Approaches

By Mikel Galar, Alberto Fernández, Edurne Barrenechea Tartas, Humberto Bustince, Francisco Herrera. 2012

1,775 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering, Media Technology, Optimization & Training, Evaluation & Benchmarks

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 Electrical and Electronic Engineering

In Media Technology

In Optimization & Training

In Evaluation & Benchmarks

Citations in the corpus

Prioritized Experience Replay (2015, neutral): “In supervised learning, there are numerous techniques to deal with imbalanced datasets when class identities are known, including re-sampling, under-sampling and over-sampling techniques, possibly combined with ensemble methods (for a review, see Galar et al., 2012).”
SMOTE for Learning from Imbalanced Data: Progress and Challenges, Marking the 15-year Anniversary (2018, background): “…point out that we may find several studies that show a good behavior for the undersampling-based approaches in synergy with ensemble learning (Khoshgoftaar, Hulse, & Napolitano, 2011; Galar et al., 2012; Blaszczynski & Stefanowski, 2015; Galar, Fernández, Barrenechea, Bustince, & Herrera, 2016).”
A comprehensive data level analysis for cancer diagnosis on imbalanced data (2019, methodology): “Researchers are interested in the problems that are created by imbalanced data classification in the real world [8,10,15–17].”
A review of epileptic seizure detection using machine learning classifiers (2020, background): “Class imbalance is one of the serious problems [140] in machine learning and the majority is seen in medical datasets [141], particularly in EEG signals.”
Automatic Denoising of Functional MRI Data: Combining Independent Component Analysis and Hierarchical Fusion of Classifiers (2014, background): “In order to compensate for one classi ﬁ er's weakness through other classi ﬁ ers' strengths, ensemble learning (also known as classi ﬁ er fusion) has been proposed — learners that combine multiple classi ﬁ ers (Galar et al., 2012; Wolpert, 1992).”
AdaFair: Cumulative Fairness Adaptive Boosting (2019, background): “n inherited problem of fairness - as we will see in our experiments, many of the datasets exhibit high class imbalance (c.f., Table 1) and therefore, tackling fairness requires also tackling imbalance[13]. Our proposed approach, AdaFair , overcomes this issue and achieves fairness while preserving h”
A Systematic Review on Imbalanced Data Challenges in Machine Learning (2019, background): “Comparative analysis of various ensemble learning algorithms and concluded that RUS-Boost algorithm yields good performance and least complex among all (Galar et al. 2012).”
Class Imbalance Problem in Data Mining Review (2013, background): “Similar situations are observed in other areas, such as detecting fraud in banking operations, detecting network intrusions [10], managing risk and predicting failures of technical”
Machine Learning in Python: Main developments and technology trends in data science, machine learning, and artificial intelligence (2020, methodology): “While the researchers use AdaBoost in this approach, potential augmentations of this method may involve other ensembling techniques.”
Predictive soil parent material mapping at a regional-scale: a Random Forest approach. (2014, background): “We recognize that the potential problem thatmay arise with the use of the area-weighted sampling approach is that such an approachwould lead to an unbalanced training dataset – a common problem for various machine-learning approaches (Galar et al., 2011; Van Hulse and Khoshgoftaar, 2009; Van Hulse e”
A review of microarray datasets and applied feature selection methods (2014, background): “In recent years, ensemble of classifiers has arisen as a possible solution to the class imbalance problem, attracting great interest among researchers [41,42], in several cases combined with preprocessing techniques such as SMOTE.”