SMOTE: Synthetic Minority Over-sampling Technique
Kevin W. BowyerNitesh V. ChawlaLawrence O. HallW. Philip KegelmeyerPublished in: CoRR (2011)
Keyphrases
- class distribution
- minority class
- class imbalance
- class imbalanced
- imbalanced data
- majority class
- imbalanced data sets
- training data
- highly imbalanced
- classification error
- cost sensitive learning
- imbalanced datasets
- cost sensitive
- support vector machine
- nearest neighbour
- test set
- decision boundary
- training set
- ensemble learning
- training dataset
- training examples
- misclassification costs
- rare events
- highly skewed
- real images are presented
- active learning
- neural network
- sampling methods
- real world
- original data
- training samples
- unlabeled data
- base classifiers
- ensemble methods
- multi class
- data sets