Analysis of preprocessing vs. cost-sensitive learning for imbalanced classification. Open problems on intrinsic data characteristics

By Victoria López, Alberto Fernández, Jose Garcia Moreno-Torres, Francisco Herrera. Expert systems with applications, 2012

312 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Electrical and Electronic Engineering

View Paper: https://www.semanticscholar.org/paper/fcbcc04418aa2c701008ee049f0915d5bdf12070

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

Hybrid case-based reasoning system by cost-sensitive neural network for classification (2016, background): “The cost-sensitive learning assumes higher misclassiﬁcation cost with samples in the minority class and hence affords to minimize high cost errors (López et al. 2012).”
How Complex Is Your Classification Problem? (2018, methodology): “In Luengo and Herrera (2015), a general automatic method for extracting the domains of competence of any ML classifier is proposed. This is done by monitoring the values of the data complexity measures and relating them to the difference in the training and testing accuracies of the classifiers. Rul”
A survey on addressing high-class imbalance in big data (2018, background): “[37] conducted a study on class imbalance with a comparison of costsensitive learning with over-sampling, and it was determined that neither technique outperformed the other.”
Fraud detection system: A survey (2016, methodology): “7 shows the balanced approaches and techniques (López et al., 2012).”
Classification with class imbalance problem: A review (2015, background): “A recent study from [90] revealed that cost-sensitive learning gives equal performance with oversampling methods and there is no difference between both strategies.”
A review of artificial neural network models for ambient air pollution prediction (2019, methodology): “, 2009), algorithm modifications (López et al., 2012), and the Synthetic Minority OverSampling Technique (SMOTE) (Chawla et al.”
A Multivariate Approach for Patient-Specific EEG Seizure Detection Using Empirical Wavelet Transform (2017, methodology): “In this study, we have applied the synthetic minority oversampling technique (SMOTE) [40] to resolve the class imbalance problem of seizure and seizure-free epochs, which neither exaggerate the receiver operating characteristic (ROC) curve of the extracted features nor cause any overfitting problem ”
Cross-Validation for Imbalanced Datasets: Avoiding Overoptimistic and Overfitting Approaches [Research Frontier] (2018, background): “Prediction models built from imbalanced datasets are most often biased towards the majority concept, which is especially critical when there is a higher cost of misclassifying the minority examples, such as diagnosing rare diseases [2].”
Cost-sensitive decision tree ensembles for effective imbalanced classification (2014, unknown): “finding since setting an appropriate cost matrix is a difficult task and in most cases there is no direct information on how to do it [33].”