CSMOTE: Contrastive Synthetic Minority Oversampling for Imbalanced Time Series Classification.
Pin LiuXiaohui GuoRui WangPengpeng ChenTianyu WoXudong LiuPublished in: ICONIP (5) (2021)
Keyphrases
- minority class
- sampling methods
- class imbalance
- majority class
- class distribution
- imbalanced data
- class imbalanced
- imbalanced datasets
- imbalanced data sets
- random sampling
- classification error
- nearest neighbour
- support vector machine
- active learning
- cost sensitive learning
- rare events
- imbalanced class distribution
- original data
- training dataset
- real world
- cost sensitive
- decision boundary
- sampling algorithm
- data sets
- training set
- single class
- database
- feature selection
- training data
- ensemble learning
- binary classification problems
- dynamic time warping
- nearest neighbor