Undersampled Majority Class Ensemble for highly imbalanced binary classification.
Pawel KsieniewiczPublished in: LIDTA@ECML/PKDD (2018)
Keyphrases
- binary classification
- class distribution
- class imbalance
- base classifiers
- cost sensitive
- imbalanced data
- ensemble methods
- minority class
- training set
- multi class
- training data
- misclassification costs
- ensemble learning
- test set
- cost sensitive learning
- binary classifiers
- generalization error
- benchmark datasets
- random forest
- prediction accuracy
- classification error
- test data
- ensemble classifier
- training samples
- sampling methods
- learning problems
- multi label
- naive bayes
- decision trees
- class labels
- supervised learning
- support vector machine
- active learning
- learning algorithm
- concept drift
- base learners
- data sets
- machine learning methods
- classification algorithm