BBW: a batch balance wrapper for training deep neural networks on extremely imbalanced datasets with few minority samples.
Jingzhao HuHao ZhangYang LiuRichard F. E. SutcliffeJun FengPublished in: Appl. Intell. (2022)
Keyphrases
- imbalanced datasets
- minority class
- majority class
- class distribution
- training dataset
- neural network
- training set
- class imbalance
- sampling methods
- training samples
- training process
- learning from imbalanced data
- nearest neighbour
- test set
- training data
- cost sensitive learning
- imbalanced data
- training examples
- support vector machine
- classification error
- feature selection
- decision boundary
- imbalanced class distribution
- original data
- ensemble methods
- active learning
- data sets
- classification accuracy
- rule extraction
- ensemble learning
- black box
- test data
- nearest neighbor
- misclassification costs
- cost sensitive
- back propagation
- base learners
- support vector
- support vectors
- artificial neural networks
- decision trees
- unlabeled data