Handling Class Imbalance in Low-Resource Dialogue Systems by Combining Few-Shot Classification and Interpolation.
Vishal SunderEric Fosler-LussierPublished in: ICASSP (2021)
Keyphrases
- class imbalance
- dialogue system
- class distribution
- cost sensitive
- cost sensitive learning
- active learning
- high dimensionality
- imbalanced datasets
- majority class
- imbalanced data sets
- sampling methods
- feature selection
- concept drift
- spoken dialogue systems
- support vector machine
- minority class
- feature extraction
- imbalanced data
- class imbalanced
- machine learning
- data sets
- decision trees
- training samples
- text classification
- dimensionality reduction
- pattern recognition
- high dimensional
- classification accuracy
- learning algorithm
- support vector
- natural language
- training set
- class labels
- feature set
- test set
- intelligent agents
- machine learning methods