Analyzing PETs on Imbalanced Datasets When Training and Testing Class Distributions Differ.
David A. CieslakNitesh V. ChawlaPublished in: PAKDD (2008)
Keyphrases
- imbalanced datasets
- class distribution
- test set
- training set
- class imbalance
- training examples
- training samples
- highly skewed
- training data
- cost sensitive learning
- test data
- training dataset
- imbalanced class distribution
- misclassification costs
- cost sensitive
- rare class
- error rate
- imbalanced data
- test cases
- supervised learning
- decision trees
- majority class
- unlabeled data
- ensemble methods
- concept drift
- sampling methods
- classification accuracy
- minority class
- pairwise
- active learning
- data sets
- feature selection algorithms
- decision boundary
- multi class
- semi supervised
- classification error
- small number
- svm classifier
- class labels
- labeled data