Learning When Data Sets are Imbalanced and When Costs are Unequal and Unknown

By M. Maloof. 2003

497 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.

Citations in the corpus

Interpretable classification models for recidivism prediction (2015, methodology): “A detailed discussion of ROC analysis in recidivism prediction can be found in the work of Maloof (2003).”
A systematic study of the class imbalance problem in convolutional neural networks (2017, methodology): “They include adjusting the decision threshold [55], considering cost-sensitive impurity function in the split criterion or applying cost-sensitive pruning schemes [60].”
Data Mining for Imbalanced Datasets: An Overview (2005, background): “can be decisive for the practitioner (Chawla, 2003; Maloof, 2003).”
ADASYN: Adaptive synthetic sampling approach for imbalanced learning (2008, methodology): “We use ROC based evaluation metrics because under the imbalanced learning condition, traditional overall classification accuracy may not be able to provide a comprehensive assessment of the observed learning algorithm [17] [31] [32] [33] [6] [34] [16].”
Editorial: special issue on learning from imbalanced data sets (2004, background): “First, a hint at the fact that research on the class imbalance problem is starting to mature is the fact that a big proportion of the 2003 workshop was occupied by the comparison of previously proposed schemes for dealing with the class imbalanced problem [9][14][37].”
Feature-Wise Bias Amplification (2018, background): “There are a myriad of empirical investigations of the effects of class imbalance in machine learning and different ways of mitigating these effects (Maloof, 2003; Chawla, 2005; Mazurowski et al., 2008; Oommen et al., 2011; Wallace et al., 2011b).”
Learning from imbalanced data sets with boosting and data generation: the DataBoost-IM approach (2004, methodology): “Following [3, 4, 5, 7], we employ the overall accuracy, G-Mean [5] and F-Measures [14] metrics to evaluate our DataBoost-IM method.”
A cost-sensitive decision tree approach for fraud detection (2013, background): “According to some studies, oversampling is effective in learning with imbalanced data sets (Japkowicz & Stephen, 2002; Japkowicz et al., 2000; Maloof, 2003).”
Joint Learning of Domain Classification and Out-of-Domain Detection with Dynamic Class Weighting for Satisficing False Acceptance Rates (2018, background): “This is relevant to addressing class imbalance or unequal class cost cases in classification, where techniques such as oversampling, undersampling, SMOTE, class weighting, and thresholdmoving are commonly used [15, 16, 17, 18, 19].”
On the Class Imbalance Problem (2008, result): “various aspects, was present in half of the papers and was the most debated issue, even though [5] shows that sampling has the same result as moving the decision threshold or adjusting the cost matrix (a result known since 1984 [6]).”
Training deep neural networks on imbalanced data sets (2016, background): “Unfortunately, in many cases, an explicit description of the cost is hard to define, instead, only an informal assertion is known like “the cost of misclassification of minority samples is higher than the contrary situation” [14].”
Classification with class imbalance problem: A review (2015, background): “Another study from [62] has demonstrated that sampling has the equivalent output to moving the decision threshold or modifying the cost matrix.”
Experimental perspectives on learning from imbalanced data (2007, unknown): “, (Drummond & Holte, 2003), (Kubat & Matwin, 1997), (Maloof, 2003), (Japkowicz, 2000), (Weiss & Provost, 2003), (Chawla et al.”
Does cost-sensitive learning beat sampling for classifying rare classes? (2005, background): “Maloof [12] also compared cost-sensitive learning to sampling but found that costsensitive learning, up-sampling and down-sampling performed nearly identically.”