SMOTE: Synthetic Minority Over-sampling Technique.
Nitesh V. ChawlaKevin W. BowyerLawrence O. HallW. Philip KegelmeyerPublished in: J. Artif. Intell. Res. (2002)
Keyphrases
- class distribution
- minority class
- class imbalance
- class imbalanced
- majority class
- imbalanced data
- imbalanced data sets
- highly imbalanced
- training set
- classification error
- decision boundary
- cost sensitive
- training data
- nearest neighbour
- original data
- test set
- rare events
- imbalanced datasets
- cost sensitive learning
- sampling methods
- misclassification costs
- real world
- support vector machine
- training dataset
- real images are presented
- feature selection
- ensemble learning
- base classifiers
- linear regression
- training samples
- feature space
- concept drift
- ensemble methods
- highly skewed
- unlabeled data