SPO: Structure Preserving Oversampling for Imbalanced Time Series Classification.
Hong CaoXiaoli LiDavid Yew-Kwong WoonSee-Kiong NgPublished in: ICDM (2011)
Keyphrases
- pairwise
- structure preserving
- class imbalance
- sampling methods
- minority class
- imbalanced datasets
- class distribution
- imbalanced data
- active learning
- multi class
- cost sensitive
- cost sensitive learning
- binary classification problems
- high dimensionality
- binary classification
- feature selection
- imbalanced class distribution
- random sampling
- dimensionality reduction
- imbalanced data sets
- multi class problems
- similarity measure
- rare events
- class imbalanced
- single class
- dynamic time warping
- classification error
- concept drift
- support vector machine
- pattern recognition
- misclassification costs
- training dataset
- ensemble learning
- prediction accuracy
- training data
- machine learning
- neural network