RUSBoost: A Hybrid Approach to Alleviating Class Imbalance

By Chris Seiffert, Taghi M. Khoshgoftaar, Jason Van Hulse, Amri Napolitano. IEEE transactions on systems, man and cybernetics. Part A. Systems and humans, 2010

1,761 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/825d543633983a2f64b6e3515c7f715378622837

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

Citations in the corpus

Classification with class imbalance problem: A review (2015, neutral): “AUC results revealed that RUSBoost, SMOTEBagging and UnderBagging return better classification compared to other ensemble algorithms particularly RUSBoost, being the least computational complex among other methods.”
Dynamically Weighted Balanced Loss: Class Imbalanced Learning and Confidence Calibration of Deep Neural Networks (2021, methodology): “Alternatively, hybrid ensemble methods that combine sampling and boosting methods [29], [30] have also been proposed in the past literature.”
An insight into classification with imbalanced data: Empirical results and current trends on using data intrinsic characteristics (2013, methodology): “In recent years, ensembles of classifiers have arisen as a possible solution to the class imbalance problem [77,85,112,117,127,131].”
Assessment of algorithms for mitosis detection in breast cancer histopathology images (2014, positive): “Since the number of non-mitosis candidate objects was disproportionately higher compared to mitosis candidate objects, the classification was performed with the RUSBoost classifier (Seiffert et al., 2010), which combines under-sampling and boosting to handle the class imbalance problem.”
Detecting Pneumatic Failures on Temporary Immersion Bioreactors (2016, methodology): “RUSBoost is a decision tree ensemble that uses a boosting algorithm [26].”
Comparing Boosting and Bagging Techniques With Noisy and Imbalanced Data (2011, methodology): “In addition to data sampling, various algorithms derived from boosting methods have been proposed for handling class imbalance (or cost-sensitive problems), including AdaCost [20], CSB0, CSB1, and CSB2 [21], AdaC1, AdaC2, and AdaC3 [22], RareBoost [23], SMOTEBoost [10], and RUSBoost [11].”
A novel ensemble method for classifying imbalanced data (2015, methodology): “Furthermore, Bagging [27], Boosting [28], and four popular Bagging and Boosting based ensemble methods EasyEnsemble [19], RUSBoost [16], SMOTEBoost [15] and UnderBagging [39] were also compared in the experiment.”
EUSBoost: Enhancing ensembles for highly imbalanced data-sets by evolutionary undersampling (2013, methodology): “The main contributions of this paper can be summarized as follows: We present a novel ensemble method based on Boosting in combination with EUS (instead of random undersampling [47] or SMOTE [12]).”
A survey on addressing high-class imbalance in big data (2018, methodology): “Adaptive Boosting (or AdaBoost) takes an iterative Boosting approach with the general goal of improving the classification performance of weak learners; some of its variants include RUSBoost [24] and ROSBoost [42].”
Multiple instance learning: A survey of problem characteristics and applications (2016, methodology): “There are external methods like SMOTE [156] and RUSBoost [157] that necessitate accurate labels to perform over or under sampling.”
Boosting methods for multi-class imbalanced data classification: an experimental review (2020, background): “It supports both binary and multiclass problems [54].”
Deep Learning Based Vulnerability Detection: Are We There Yet? (2020, background): “Existing approaches do not alleviate the class imbalance problem [10] of real-world vul-”
A Survey of Predictive Modeling on Imbalanced Domains (2016, methodology): “Namely, it was combined with boosting [Seiffert et al. 2010] and bagging [Chang et al. 2003; Tao et al. 2006; Wang and Yao 2009; Wallace et al. 2011] and was applied to both classes in random forests in a method named Balanced Random Forest (BRF) [Chen et al. 2004].”
Calibrated asymmetric surrogate losses (2012, background): “Examples include boosting algorithms [5, 6, 7, 9, 13, 16, 20, 29, 33, 34, 36, 38, 39, 40], support vector machines [1, 4, 8, 10, 12, 18, 22, 37, 41, 42], and the perceptron [11].”